{ "aboutHtml": "

rerank-2.5 adalah reranker Voyage AI yang mengikuti instruksi, dirancang untuk menyempurnakan kumpulan kandidat pencarian. Model ini memeriksa kueri dan setiap dokumen secara bersamaan sebagai cross-encoder, lalu mengembalikan urutan relevansi. Rilis pembuat menekankan kendali bahasa alami atas hal yang dianggap relevan, pengambilan informasi multibahasa, dan konteks kueri-dokumen yang lebih panjang daripada edisi rerank-2 sebelumnya.

Berikan kandidat yang perlu dibandingkan kepada model

Masukannya adalah kueri dan daftar dokumen teks kandidat. Kandidat dapat berupa bagian teks yang dipilih pencarian embedding atau pencarian leksikal tahap pertama. Reranker mengevaluasi teks yang diberikan, bukan membuat embedding korpus, sehingga tugas praktisnya adalah menyempurnakan kandidat yang sudah ditemukan pencarian awal.

Pertahankan urutan kandidat asli dan pengidentifikasi sumber saat menyiapkan daftar tersebut. Indeks yang dikembalikan merujuk pada posisi masukan itu. Untuk tugas RAG, berikan bagian teks yang mungkin benar-benar digunakan tahap jawaban, termasuk judul atau kondisi yang diperlukan untuk menafsirkannya. Dokumen relevan yang tidak ada dalam kumpulan kandidat yang dikirim tidak dapat muncul sebagai hasil yang baru ditemukan melalui pemeringkatan ulang.

Tentukan pengertian relevansi yang dituju

Instruksi dapat ditambahkan sesudah atau sebelum kueri. Rilis tersebut menjelaskan tiga peran yang berguna: menekankan bagian tertentu dari dokumen, mengutamakan jenis sumber, dan memperjelas kueri dengan konteks aplikasi. Ketiganya mengubah tugas relevansi yang disajikan kepada model.

Untuk pencarian teknis, contoh instruksinya adalah “Prioritaskan bagian teks yang menjelaskan perilaku kegagalan dan pemulihan; gambaran pengantar bersifat sekunder.” Untuk kumpulan riset, tentukan apakah judul, metode, atau hasil yang paling penting. Untuk istilah yang ambigu, nyatakan bidang subjek pencarian. Jaga pertanyaan dan instruksi tetap selaras agar yang satu tidak meminta topik luas sementara yang lain diam-diam mengubah tugas.

Preferensi ini memandu pemberian skor relevansi. Jika aplikasi memiliki persyaratan kelayakan ketat—seperti versi sumber, tanggal, atau cakupan akses—siapkan kumpulan kandidat yang memenuhi syarat secara terpisah. Instruksi relevansi model dan pembatasan koleksi deterministik melayani tujuan berbeda.

Keluaran berperingkat dan penggunaan hasil berikutnya

Hasil diurutkan berdasarkan relevansi menurun dan mencakup skor relevansi floating-point serta indeks masukan asli kandidat. Opsi top-k memilih jumlah hasil teratas yang dikembalikan; tanpa opsi tersebut, spesifikasi pembuat mengembalikan hasil untuk semua kandidat yang dikirim. Pengaturan return-documents opsional menyertakan teks sumber yang bersesuaian; jika tidak, respons memuat indeks dan skor.

Gunakan urutan tersebut untuk memilih bagian teks yang akan ditampilkan atau digunakan pada tahap jawaban berikutnya, dengan menjaga identitas sumbernya tetap utuh. Periksa teks aktual ketika mengevaluasi peringkat: skor adalah ukuran relevansi terhadap kueri yang dikirim, bukan jawaban yang dihasilkan atau verifikasi bahwa klaim dokumen benar. Ini membuat model berguna untuk menyempurnakan materi pencarian teknis, hukum, keuangan, medis, percakapan, dan multibahasa tanpa mengubah pemeringkatan menjadi kesimpulan bidang tersebut.

Konteks, anggaran total token, dan truncation

Batas kueri yang didokumentasikan adalah 8.000 token, termasuk instruksi yang ditempatkan dalam kueri. Kueri ditambah satu dokumen kandidat mana pun harus muat dalam 32.000 token, dan daftar dokumen dapat memuat paling banyak 1.000 item. Alokasi total adalah 600.000 token, dihitung sebagai token kueri dikalikan jumlah dokumen, ditambah seluruh token dokumen. Karena itu, instruksi panjang dihitung berulang di seluruh pasangan kandidat.

Jumlah top-k keluaran tidak mengubah jumlah kandidat dalam perhitungan token masukan tersebut. Truncation default memendekkan kueri atau dokumen yang terlalu panjang agar memenuhi batas konteks; menonaktifkannya membuat kueri atau pasangan yang terlalu panjang mengembalikan kesalahan. Siapkan bagian teks yang memuat jawaban secara sengaja dan jaga agar kondisi yang relevan tetap berada dalam teks yang akan dilihat model. Evaluasi urutan yang disempurnakan pada pertanyaan dan instruksi representatif, lalu teruskan bagian teks asli yang dipilih ke tahap berikutnya.

", "aboutText": "rerank-2.5 adalah reranker Voyage AI yang mengikuti instruksi, dirancang untuk menyempurnakan kumpulan kandidat pencarian. Model ini memeriksa kueri dan setiap dokumen secara bersamaan sebagai cross-encoder, lalu mengembalikan urutan relevansi. Rilis pembuat menekankan kendali bahasa alami atas hal yang dianggap relevan, pengambilan informasi multibahasa, dan konteks kueri-dokumen yang lebih panjang daripada edisi rerank-2 sebelumnya.\n\n### Berikan kandidat yang perlu dibandingkan kepada model\n\nMasukannya adalah kueri dan daftar dokumen teks kandidat. Kandidat dapat berupa bagian teks yang dipilih pencarian embedding atau pencarian leksikal tahap pertama. Reranker mengevaluasi teks yang diberikan, bukan membuat embedding korpus, sehingga tugas praktisnya adalah menyempurnakan kandidat yang sudah ditemukan pencarian awal.\n\nPertahankan urutan kandidat asli dan pengidentifikasi sumber saat menyiapkan daftar tersebut. Indeks yang dikembalikan merujuk pada posisi masukan itu. Untuk tugas RAG, berikan bagian teks yang mungkin benar-benar digunakan tahap jawaban, termasuk judul atau kondisi yang diperlukan untuk menafsirkannya. Dokumen relevan yang tidak ada dalam kumpulan kandidat yang dikirim tidak dapat muncul sebagai hasil yang baru ditemukan melalui pemeringkatan ulang.\n\n### Tentukan pengertian relevansi yang dituju\n\nInstruksi dapat ditambahkan sesudah atau sebelum kueri. Rilis tersebut menjelaskan tiga peran yang berguna: menekankan bagian tertentu dari dokumen, mengutamakan jenis sumber, dan memperjelas kueri dengan konteks aplikasi. Ketiganya mengubah tugas relevansi yang disajikan kepada model.\n\nUntuk pencarian teknis, contoh instruksinya adalah “Prioritaskan bagian teks yang menjelaskan perilaku kegagalan dan pemulihan; gambaran pengantar bersifat sekunder.” Untuk kumpulan riset, tentukan apakah judul, metode, atau hasil yang paling penting. Untuk istilah yang ambigu, nyatakan bidang subjek pencarian. Jaga pertanyaan dan instruksi tetap selaras agar yang satu tidak meminta topik luas sementara yang lain diam-diam mengubah tugas.\n\nPreferensi ini memandu pemberian skor relevansi. Jika aplikasi memiliki persyaratan kelayakan ketat—seperti versi sumber, tanggal, atau cakupan akses—siapkan kumpulan kandidat yang memenuhi syarat secara terpisah. Instruksi relevansi model dan pembatasan koleksi deterministik melayani tujuan berbeda.\n\n### Keluaran berperingkat dan penggunaan hasil berikutnya\n\nHasil diurutkan berdasarkan relevansi menurun dan mencakup skor relevansi floating-point serta indeks masukan asli kandidat. Opsi top-k memilih jumlah hasil teratas yang dikembalikan; tanpa opsi tersebut, spesifikasi pembuat mengembalikan hasil untuk semua kandidat yang dikirim. Pengaturan return-documents opsional menyertakan teks sumber yang bersesuaian; jika tidak, respons memuat indeks dan skor.\n\nGunakan urutan tersebut untuk memilih bagian teks yang akan ditampilkan atau digunakan pada tahap jawaban berikutnya, dengan menjaga identitas sumbernya tetap utuh. Periksa teks aktual ketika mengevaluasi peringkat: skor adalah ukuran relevansi terhadap kueri yang dikirim, bukan jawaban yang dihasilkan atau verifikasi bahwa klaim dokumen benar. Ini membuat model berguna untuk menyempurnakan materi pencarian teknis, hukum, keuangan, medis, percakapan, dan multibahasa tanpa mengubah pemeringkatan menjadi kesimpulan bidang tersebut.\n\n### Konteks, anggaran total token, dan truncation\n\nBatas kueri yang didokumentasikan adalah 8.000 token, termasuk instruksi yang ditempatkan dalam kueri. Kueri ditambah satu dokumen kandidat mana pun harus muat dalam 32.000 token, dan daftar dokumen dapat memuat paling banyak 1.000 item. Alokasi total adalah 600.000 token, dihitung sebagai token kueri dikalikan jumlah dokumen, ditambah seluruh token dokumen. Karena itu, instruksi panjang dihitung berulang di seluruh pasangan kandidat.\n\nJumlah top-k keluaran tidak mengubah jumlah kandidat dalam perhitungan token masukan tersebut. Truncation default memendekkan kueri atau dokumen yang terlalu panjang agar memenuhi batas konteks; menonaktifkannya membuat kueri atau pasangan yang terlalu panjang mengembalikan kesalahan. Siapkan bagian teks yang memuat jawaban secara sengaja dan jaga agar kondisi yang relevan tetap berada dalam teks yang akan dilihat model. Evaluasi urutan yang disempurnakan pada pertanyaan dan instruksi representatif, lalu teruskan bagian teks asli yang dipilih ke tahap berikutnya.", "editorialOverlay": { "copy_fields": null, "copy_sha256": null, "copy_source": null, "html_sha256": "b93544bf73442b4418f0c0731b92e5a5c195df5d3b49b9752847632181797fe5", "module_sha256": "ee9dc1c50b1e6a8fb7fa062030fcb4646b9aaaeb08be513b723a5a92a7290f06", "raw_model_unchanged": true, "schema": "ailoot-native-model-editorial-v1" }, "locale": "id", "methods": [ { "html": "
•

VoyageAI by MongoDB: rerank-2.5

VoyageAI by MongoDB /voyageai/rerank-2.5

Mulai

rerank-2.5 mengurutkan ulang kumpulan kandidat teks yang diberikan berdasarkan relevansi terhadap kueri dan instruksi bahasa alami opsional. Ini adalah reranker multibahasa yang berorientasi pada kualitas, dengan konteks kueri-dokumen 32.000 token dan skor berperingkat yang terkait dengan indeks kandidat asli.

Masukan yang diterima
Teks
Hasil
Hasil berperingkat
Tentang model

Deskripsi ini mencakup kemampuan model. Masukan, keluaran, dan opsi yang tersedia di Ailoot tercantum di bawah untuk metode yang dipilih.

rerank-2.5 adalah reranker Voyage AI yang mengikuti instruksi, dirancang untuk menyempurnakan kumpulan kandidat pencarian. Model ini memeriksa kueri dan setiap dokumen secara bersamaan sebagai cross-encoder, lalu mengembalikan urutan relevansi. Rilis pembuat menekankan kendali bahasa alami atas hal yang dianggap relevan, pengambilan informasi multibahasa, dan konteks kueri-dokumen yang lebih panjang daripada edisi rerank-2 sebelumnya.

Berikan kandidat yang perlu dibandingkan kepada model

Masukannya adalah kueri dan daftar dokumen teks kandidat. Kandidat dapat berupa bagian teks yang dipilih pencarian embedding atau pencarian leksikal tahap pertama. Reranker mengevaluasi teks yang diberikan, bukan membuat embedding korpus, sehingga tugas praktisnya adalah menyempurnakan kandidat yang sudah ditemukan pencarian awal.

Pertahankan urutan kandidat asli dan pengidentifikasi sumber saat menyiapkan daftar tersebut. Indeks yang dikembalikan merujuk pada posisi masukan itu. Untuk tugas RAG, berikan bagian teks yang mungkin benar-benar digunakan tahap jawaban, termasuk judul atau kondisi yang diperlukan untuk menafsirkannya. Dokumen relevan yang tidak ada dalam kumpulan kandidat yang dikirim tidak dapat muncul sebagai hasil yang baru ditemukan melalui pemeringkatan ulang.

Tentukan pengertian relevansi yang dituju

Instruksi dapat ditambahkan sesudah atau sebelum kueri. Rilis tersebut menjelaskan tiga peran yang berguna: menekankan bagian tertentu dari dokumen, mengutamakan jenis sumber, dan memperjelas kueri dengan konteks aplikasi. Ketiganya mengubah tugas relevansi yang disajikan kepada model.

Untuk pencarian teknis, contoh instruksinya adalah “Prioritaskan bagian teks yang menjelaskan perilaku kegagalan dan pemulihan; gambaran pengantar bersifat sekunder.” Untuk kumpulan riset, tentukan apakah judul, metode, atau hasil yang paling penting. Untuk istilah yang ambigu, nyatakan bidang subjek pencarian. Jaga pertanyaan dan instruksi tetap selaras agar yang satu tidak meminta topik luas sementara yang lain diam-diam mengubah tugas.

Preferensi ini memandu pemberian skor relevansi. Jika aplikasi memiliki persyaratan kelayakan ketat—seperti versi sumber, tanggal, atau cakupan akses—siapkan kumpulan kandidat yang memenuhi syarat secara terpisah. Instruksi relevansi model dan pembatasan koleksi deterministik melayani tujuan berbeda.

Keluaran berperingkat dan penggunaan hasil berikutnya

Hasil diurutkan berdasarkan relevansi menurun dan mencakup skor relevansi floating-point serta indeks masukan asli kandidat. Opsi top-k memilih jumlah hasil teratas yang dikembalikan; tanpa opsi tersebut, spesifikasi pembuat mengembalikan hasil untuk semua kandidat yang dikirim. Pengaturan return-documents opsional menyertakan teks sumber yang bersesuaian; jika tidak, respons memuat indeks dan skor.

Gunakan urutan tersebut untuk memilih bagian teks yang akan ditampilkan atau digunakan pada tahap jawaban berikutnya, dengan menjaga identitas sumbernya tetap utuh. Periksa teks aktual ketika mengevaluasi peringkat: skor adalah ukuran relevansi terhadap kueri yang dikirim, bukan jawaban yang dihasilkan atau verifikasi bahwa klaim dokumen benar. Ini membuat model berguna untuk menyempurnakan materi pencarian teknis, hukum, keuangan, medis, percakapan, dan multibahasa tanpa mengubah pemeringkatan menjadi kesimpulan bidang tersebut.

Konteks, anggaran total token, dan truncation

Batas kueri yang didokumentasikan adalah 8.000 token, termasuk instruksi yang ditempatkan dalam kueri. Kueri ditambah satu dokumen kandidat mana pun harus muat dalam 32.000 token, dan daftar dokumen dapat memuat paling banyak 1.000 item. Alokasi total adalah 600.000 token, dihitung sebagai token kueri dikalikan jumlah dokumen, ditambah seluruh token dokumen. Karena itu, instruksi panjang dihitung berulang di seluruh pasangan kandidat.

Jumlah top-k keluaran tidak mengubah jumlah kandidat dalam perhitungan token masukan tersebut. Truncation default memendekkan kueri atau dokumen yang terlalu panjang agar memenuhi batas konteks; menonaktifkannya membuat kueri atau pasangan yang terlalu panjang mengembalikan kesalahan. Siapkan bagian teks yang memuat jawaban secara sengaja dan jaga agar kondisi yang relevan tetap berada dalam teks yang akan dilihat model. Evaluasi urutan yang disempurnakan pada pertanyaan dan instruksi representatif, lalu teruskan bagian teks asli yang dipilih ke tahap berikutnya.

Tersedia di Ailoot · masukan & metode

Ailoot Ultra · Pemeringkatan ulang POST /v1/rerank
Masukan
text
Keluaran
rerank
Format keluaran
json
KolomJenisPersyaratanNilai & batasanBawaan
KolomdocumentsJenisarrayPersyaratanWajibNilai & batasanBawaan—
KolommodelJenisstringPersyaratanWajibNilai & batasan
voyageai/rerank-2.5
Bawaan—
KolomproviderJenisschemaPersyaratanOpsionalNilai & batasanBawaan—
KolomqueryJenisstringPersyaratanWajibNilai & batasanBawaan—
Kolomsession_idJenisstringPersyaratanOpsionalNilai & batasanBawaan—
Kolomtop_nJenisintegerPersyaratanOpsionalNilai & batasan
Minimum: 1
Bawaan—
KolomtraceJenisschemaPersyaratanOpsionalNilai & batasanBawaan—
KolomuserJenisstringPersyaratanOpsionalNilai & batasanBawaan—
Kolomdocuments[].textJenisstringPersyaratanOpsionalNilai & batasanBawaan—
Kolomdocuments[].imageJenisstringPersyaratanOpsionalNilai & batasanBawaan—
Kombinasi bahan masukan

text_rerank · Wajib: query, documents array with minItems1

Parameter metode dan siklus hidup
ParameterKontrak
Parametercontent_typeKontrakapplication/json
ParameterrequiredKontrakmodel, query, documents
Parametermaterial_typesKontraktext
Parameteroptional_controlsKontrakprovider, session_id, top_n, trace, user
Parametermultimodal_input_bindingKontraknull

Siklus hidup: synchronous

Kirim ID model yang persis dan simpan respons native secara utuh.

Paket

Referensi tarif tidak menambah opsi permintaan. Parameter yang didukung tercantum pada metode yang dipilih.

Mode Fast
Tidak ditentukan Saat mode fast dipilih
Mode Flex
Tidak ditentukan Saat mode flex dipilih
Upaya penalaran maksimum
Tidak ditentukan Saat upaya penalaran maksimum menggunakan penyesuaian default

Tarif Ailoot

Harga menurut konfigurasi

Cocokkan setiap biaya dengan kondisi yang ditampilkan. Gunakan satu tarif untuk setiap kelompok konfigurasi alternatif.

KonfigurasiJadwal tarifSatuan penagihanBiaya
KonfigurasiTarif dasar tokenJadwal tarifTarif dasarSatuan penagihan—Biaya
Masukan · Token masukan — / 1K token
Keluaran · Token keluaran — / 1K token
Masukan · Pembacaan cache — / 1K token
Masukan · Penulisan cache — / 1K token
Masukan · Penulisan cache, 1 jam — / 1K token
KonfigurasiUltraJadwal tarif—Satuan penagihan/ 1K tokenBiaya
Masukan $0.0001 / 1K token

Ailoot mencadangkan perkiraan dana sebelum permintaan. Setelah hasil diterima, tagihan akhir diselesaikan berdasarkan penggunaan yang terukur. Penyelesaian tagihan streaming menggunakan catatan penggunaan akhir. Cara kerja cadangan dan penyelesaian tagihan →

Contoh permintaan

Ailoot Ultra · Pemeringkatan ulang

{\n  "documents": [\n    "Paris is the capital of France.",\n    "Berlin is the capital of Germany."\n  ],\n  "model": "voyageai/rerank-2.5",\n  "query": "What is the capital of France?",\n  "top_n": 3\n}

Terima dan simpan hasil

Ailoot Ultra · Pemeringkatan ulang

json

Simpan semua hasil yang selesai

Simpan setiap respons lengkap, berkas, dan bagian konten yang dikembalikan.

Setiap indeks hasil aktual, relevance_score yang bernilai berhingga, dan objek dokumen asli dipertahankan. Indeks mengacu pada array dokumen asli. Seluruh JSON asli disimpan dan hanya dikembalikan setelah penyelesaian tagihan.

Jika respons asli perlu diperiksa

Simpan operasi awal dan balasan lengkap/sebagian; jangan ulangi POST.

Kolom respons
KolomJenis
KolommodelJenisstring
KolomidJenisstring
KolomproviderJenisstring
Kolomresults[].indexJenisinteger
Kolomresults[].relevance_scoreJenisnumber
Kolomresults[].document.textJenisstring
Kolomresults[].document.imageJenisstring
Kolomusage.search_unitsJenisinteger
Kolomusage.total_tokensJenisinteger
Kolomusage.costJenisnumber
Kesalahan dan langkah berikutnya
ResponsKondisiLangkah berikutnya
Respons400 · invalid_request_errorKondisiBidang wajib, representasi, atau nilai tidak validLangkah berikutnyaPerbaiki bidang permintaan yang ditunjukkan.

Alias

Alias model

Ailoot Ultra voyageai/rerank-2.5

Alias jalur metode

Gunakan jalur metode yang ditampilkan di atas.

ID pemanggil terkait

voyageai/rerank-2.5

", "method_identity": { "effective_request_model_id": "voyageai/rerank-2.5", "fmt": "rerank", "http": "POST", "path": "/v1/rerank", "request_model_id": "voyageai/rerank-2.5", "route_label": "Ultra" }, "method_key": "Ultra|rerank|/v1/rerank|voyageai%2Frerank-2.5", "ordinal": 0, "rawMethod": { "available": null, "cache_read": null, "cache_write": null, "cache_write_1h": null, "contract": { "data": { "all_results": "Every actual result index, finite relevance_score and original document object is retained. Indices bind the original document array. Entire original JSON is stored and returned only after settlement.", "errors": [ { "code": null, "condition": "Bidang wajib, representasi, atau nilai tidak valid", "http_status": 400, "recovery": "Perbaiki bidang permintaan yang ditunjukkan.", "type": "invalid_request_error" } ], "hints": [], "ignored_fields": [], "lifecycle": { "cancel": null, "idempotency_header": null, "job_retrieve": null, "mode": "synchronous", "streaming": false, "unknown": "Simpan operasi awal dan balasan lengkap/sebagian; jangan ulangi POST." }, "material_combinations": [ { "accepted": [ "string documents", "objects with text; image only for separately exact-supported multimodal model" ], "mode": "text_rerank", "required": [ "query", "documents array with minItems1" ] } ], "operation_modes": [ "rerank" ], "outputs": { "all_results": "Setiap indeks hasil aktual, relevance_score yang bernilai berhingga, dan objek dokumen asli dipertahankan. Indeks mengacu pada array dokumen asli. Seluruh JSON asli disimpan dan hanya dikembalikan setelah penyelesaian tagihan.", "family": "rerank", "fields": [ { "description": null, "name": "model", "type": "string" }, { "description": null, "name": "id", "type": "string" }, { "description": null, "name": "provider", "type": "string" }, { "description": null, "name": "results[].index", "type": "integer" }, { "description": null, "name": "results[].relevance_score", "type": "number" }, { "description": null, "name": "results[].document.text", "type": "string" }, { "description": null, "name": "results[].document.image", "type": "string" }, { "description": null, "name": "usage.search_units", "type": "integer" }, { "description": null, "name": "usage.total_tokens", "type": "integer" }, { "description": null, "name": "usage.cost", "type": "number" } ], "format": "json", "schema": { "description": "Rerank response containing ranked results", "example": { "id": "gen-rerank-1234567890-abc", "model": "cohere/rerank-v3.5", "results": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "usage": { "search_units": 1, "total_tokens": 150 } }, "properties": { "id": { "description": "Unique identifier for the rerank response (ORID format)", "example": "gen-rerank-1234567890-abc", "type": "string" }, "model": { "description": "The model used for reranking", "example": "cohere/rerank-v3.5", "type": "string" }, "provider": { "description": "The provider that served the rerank request", "example": "Cohere", "type": "string" }, "results": { "description": "List of rerank results sorted by relevance", "example": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "items": { "description": "A single rerank result", "example": { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 }, "properties": { "document": { "description": "The document object echoing the original input (text and/or image)", "properties": { "image": { "description": "The image (URL or data URI) from the original document", "example": "https://example.com/image.png", "type": "string" }, "text": { "description": "The document text", "example": "Paris is the capital of France.", "type": "string" } }, "type": "object" }, "index": { "description": "Index of the document in the original input list", "example": 0, "type": "integer" }, "relevance_score": { "description": "Relevance score of the document to the query", "example": 0.98, "format": "double", "type": "number" } }, "required": [ "index", "relevance_score", "document" ], "type": "object" }, "type": "array" }, "usage": { "description": "Usage statistics", "example": { "search_units": 1, "total_tokens": 150 }, "properties": { "cost": { "description": "Cost of the request in credits", "example": 0.001, "format": "double", "type": "number" }, "search_units": { "description": "Number of search units consumed (Cohere billing)", "example": 1, "type": "integer" }, "total_tokens": { "description": "Total number of tokens used", "example": 150, "type": "integer" } }, "type": "object" } }, "required": [ "model", "results" ], "type": "object" } }, "parameters": [ { "availability": "source_interface", "default": null, "default_known": false, "description": "The list of documents to rerank. Documents may be plain strings, or structured objects with `text` and/or `image` for multimodal models.", "enum": null, "enum_known": false, "minItems": 1, "name": "documents", "required": true, "type": "array" }, { "availability": "source_interface", "default": null, "default_known": false, "enum": [ "voyageai/rerank-2.5" ], "enum_known": true, "name": "model", "required": true, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "enum": null, "enum_known": false, "name": "provider", "required": false, "type": "schema" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "The search query to rerank documents against", "enum": null, "enum_known": false, "name": "query", "required": true, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "A unique identifier for grouping related requests (e.g., a conversation or agent workflow). Used for observability grouping in Broadcast and private logging; never sent to the provider. If provided in both the request body and the x-session-id header, the body value takes precedence. Maximum of 256 characters.", "enum": null, "enum_known": false, "maxLength": 256, "name": "session_id", "required": false, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "Number of most relevant documents to return", "enum": null, "enum_known": false, "minimum": 1, "name": "top_n", "required": false, "type": "integer" }, { "availability": "source_interface", "default": null, "default_known": false, "enum": null, "enum_known": false, "name": "trace", "required": false, "type": "schema" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "A unique identifier representing your end-user. Forwarded to Broadcast and private logging as the end-user id; never sent to the provider.", "enum": null, "enum_known": false, "maxLength": 256, "name": "user", "required": false, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "enum": null, "enum_known": false, "name": "documents[].text", "required": false, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "Official family image carrier is remote http/https URL or data:image URI; only for exact multimodal-model bindings.", "enum": null, "enum_known": false, "name": "documents[].image", "required": false, "type": "string" } ], "unsupported": [ "stream=true" ] }, "hints": [ "Kirim ID model yang persis dan simpan respons native secara utuh." ], "input": { "content_type": "application/json", "material_types": [ "text" ], "multimodal_input_binding": null, "optional_controls": [ "provider", "session_id", "top_n", "trace", "user" ], "required": [ "model", "query", "documents" ] }, "lifecycle": { "cancel": null, "idempotency_header": null, "job_retrieve": null, "mode": "synchronous", "streaming": false, "unknown": "Preserve original operation and full/partial reply; do not replay POST." }, "manufacturer_capability": { "context_length": 32000, "manufacturer_input": [ "text" ], "manufacturer_output": [ "rerank" ] }, "method": { "http": "POST", "path": "/v1/rerank" }, "method_aliases": [], "operation_modes": [ "rerank" ], "output": { "all_results": "Every actual result index, finite relevance_score and original document object is retained. Indices bind the original document array. Entire original JSON is stored and returned only after settlement.", "family": "rerank", "fields": [ { "description": null, "name": "model", "type": "string" }, { "description": null, "name": "id", "type": "string" }, { "description": null, "name": "provider", "type": "string" }, { "description": null, "name": "results[].index", "type": "integer" }, { "description": null, "name": "results[].relevance_score", "type": "number" }, { "description": null, "name": "results[].document.text", "type": "string" }, { "description": null, "name": "results[].document.image", "type": "string" }, { "description": null, "name": "usage.search_units", "type": "integer" }, { "description": null, "name": "usage.total_tokens", "type": "integer" }, { "description": null, "name": "usage.cost", "type": "number" } ], "format": "json", "schema": { "description": "Rerank response containing ranked results", "example": { "id": "gen-rerank-1234567890-abc", "model": "cohere/rerank-v3.5", "results": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "usage": { "search_units": 1, "total_tokens": 150 } }, "properties": { "id": { "description": "Unique identifier for the rerank response (ORID format)", "example": "gen-rerank-1234567890-abc", "type": "string" }, "model": { "description": "The model used for reranking", "example": "cohere/rerank-v3.5", "type": "string" }, "provider": { "description": "The provider that served the rerank request", "example": "Cohere", "type": "string" }, "results": { "description": "List of rerank results sorted by relevance", "example": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "items": { "description": "A single rerank result", "example": { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 }, "properties": { "document": { "description": "The document object echoing the original input (text and/or image)", "properties": { "image": { "description": "The image (URL or data URI) from the original document", "example": "https://example.com/image.png", "type": "string" }, "text": { "description": "The document text", "example": "Paris is the capital of France.", "type": "string" } }, "type": "object" }, "index": { "description": "Index of the document in the original input list", "example": 0, "type": "integer" }, "relevance_score": { "description": "Relevance score of the document to the query", "example": 0.98, "format": "double", "type": "number" } }, "required": [ "index", "relevance_score", "document" ], "type": "object" }, "type": "array" }, "usage": { "description": "Usage statistics", "example": { "search_units": 1, "total_tokens": 150 }, "properties": { "cost": { "description": "Cost of the request in credits", "example": 0.001, "format": "double", "type": "number" }, "search_units": { "description": "Number of search units consumed (Cohere billing)", "example": 1, "type": "integer" }, "total_tokens": { "description": "Total number of tokens used", "example": 150, "type": "integer" } }, "type": "object" } }, "required": [ "model", "results" ], "type": "object" } }, "schema": { "request": { "description": "Rerank request input", "example": { "documents": [ "Paris is the capital of France.", "Berlin is the capital of Germany." ], "model": "cohere/rerank-v3.5", "query": "What is the capital of France?", "top_n": 3 }, "properties": { "documents": { "description": "The list of documents to rerank. Documents may be plain strings, or structured objects with `text` and/or `image` for multimodal models.", "example": [ "Paris is the capital of France.", "Berlin is the capital of Germany." ], "items": { "anyOf": [ { "type": "string" }, { "description": "A structured document with optional text and/or image content. At least one of `text` or `image` must be provided.", "properties": { "image": { "description": "An image associated with the document, as a remote URL (http/https) or a base64-encoded data URI (data:image/...).", "example": "https://upload.wikimedia.org/wikipedia/commons/thumb/8/8b/Phytogenic.png", "type": "string" }, "text": { "description": "The document text", "example": "AI enables robots to perceive, plan, and act autonomously.", "type": "string" } }, "type": "object" } ], "description": "A document to rerank. Either a plain string, or a structured object with optional `text` and/or `image`." }, "minItems": 1, "type": "array" }, "model": { "enum": [ "voyageai/rerank-2.5" ], "type": "string" }, "provider": { "allOf": [ { "$ref": "#/components/schemas/ProviderPreferences" }, { "description": "Provider routing preferences for the request." } ] }, "query": { "description": "The search query to rerank documents against", "example": "What is the capital of France?", "type": "string" }, "session_id": { "description": "A unique identifier for grouping related requests (e.g., a conversation or agent workflow). Used for observability grouping in Broadcast and private logging; never sent to the provider. If provided in both the request body and the x-session-id header, the body value takes precedence. Maximum of 256 characters.", "example": "session-1234", "maxLength": 256, "type": "string" }, "top_n": { "description": "Number of most relevant documents to return", "example": 3, "minimum": 1, "type": "integer" }, "trace": { "$ref": "#/components/schemas/TraceConfig" }, "user": { "description": "A unique identifier representing your end-user. Forwarded to Broadcast and private logging as the end-user id; never sent to the provider.", "example": "user-1234", "maxLength": 256, "type": "string" } }, "required": [ "model", "query", "documents" ], "type": "object" }, "response": { "description": "Rerank response containing ranked results", "example": { "id": "gen-rerank-1234567890-abc", "model": "cohere/rerank-v3.5", "results": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "usage": { "search_units": 1, "total_tokens": 150 } }, "properties": { "id": { "description": "Unique identifier for the rerank response (ORID format)", "example": "gen-rerank-1234567890-abc", "type": "string" }, "model": { "description": "The model used for reranking", "example": "cohere/rerank-v3.5", "type": "string" }, "provider": { "description": "The provider that served the rerank request", "example": "Cohere", "type": "string" }, "results": { "description": "List of rerank results sorted by relevance", "example": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "items": { "description": "A single rerank result", "example": { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 }, "properties": { "document": { "description": "The document object echoing the original input (text and/or image)", "properties": { "image": { "description": "The image (URL or data URI) from the original document", "example": "https://example.com/image.png", "type": "string" }, "text": { "description": "The document text", "example": "Paris is the capital of France.", "type": "string" } }, "type": "object" }, "index": { "description": "Index of the document in the original input list", "example": 0, "type": "integer" }, "relevance_score": { "description": "Relevance score of the document to the query", "example": 0.98, "format": "double", "type": "number" } }, "required": [ "index", "relevance_score", "document" ], "type": "object" }, "type": "array" }, "usage": { "description": "Usage statistics", "example": { "search_units": 1, "total_tokens": 150 }, "properties": { "cost": { "description": "Cost of the request in credits", "example": 0.001, "format": "double", "type": "number" }, "search_units": { "description": "Number of search units consumed (Cohere billing)", "example": 1, "type": "integer" }, "total_tokens": { "description": "Total number of tokens used", "example": 150, "type": "integer" } }, "type": "object" } }, "required": [ "model", "results" ], "type": "object" } } }, "ctx": null, "declared_pricing_profiles": [ { "dimensions": { "regime": "base" }, "label": "Base token rates", "units": [ { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_price", "label": "Input tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "output_price", "label": "Output tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "output_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "out", "unit": "output_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_read_price", "label": "Cache read", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "cached_input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "cached_input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_price", "label": "Cache write", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "generic_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "generic_cache_creation_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_1h_price", "label": "Cache write, 1 hour", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "one_hour_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "one_hour_cache_creation_token" } ] }, { "dimensions": { "native_endpoint": "rerank", "quoted_unit": "/ 1K tokens" }, "label": "Ultra", "units": [ { "amount": "0.0001000", "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_token", "label": "input token", "native_amount_usd_per_unit": "0.0000001", "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" } ] } ], "examples": [ { "code": "{\n \"documents\": [\n \"Paris is the capital of France.\",\n \"Berlin is the capital of Germany.\"\n ],\n \"model\": \"voyageai/rerank-2.5\",\n \"query\": \"What is the capital of France?\",\n \"top_n\": 3\n}", "label": "Retained official request structure; documentary example", "language": "json" }, { "code": "set -eu\n: \"${AILOOT_API_KEY:?Set your Ailoot API key}\"\n: \"${AILOOT_REQUEST_FILE:?Set your actual JSON request file}\"\nexport AILOOT_REQUEST_FILE\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads(Path(os.environ['AILOOT_REQUEST_FILE']).read_text(encoding='utf-8'))\nif not isinstance(p,dict): raise SystemExit('Request must be a JSON object')\np['model']='voyageai/rerank-2.5'\nPath('request.json').write_text(json.dumps(p,ensure_ascii=False),encoding='utf-8')\nPY_REQUEST\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/rerank \\\n -H \"Authorization: Bearer $AILOOT_API_KEY\" -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\npython3 - <<'PY_RESULT'\nimport json\nfrom pathlib import Path\nprint(json.dumps(json.loads(Path('result.json').read_text()),ensure_ascii=False,indent=2))\nPY_RESULT\n", "label": "Use your actual request material and retain the complete result", "language": "bash", "material_policy": "Use real owner-supplied state/questions or query/documents. No invented file ID, document, image URL, job ID or returned result." } ], "family": "rerank", "fast_multiplier": null, "flex_multiplier": null, "fmt": "rerank", "http": "POST", "in": null, "input": [ "text" ], "label": "Pemeringkatan ulang", "max_out": null, "max_reasoning_effort_multiplier": null, "method_aliases": [], "native_endpoint": "rerank", "out": null, "output": [ "rerank" ], "path": "/v1/rerank", "price_conditions": [ { "label": "Fast mode", "value": "Not specified", "when": "When fast mode is selected" }, { "label": "Flex mode", "value": "Not specified", "when": "When flex mode is selected" }, { "label": "Maximum reasoning effort", "value": "Not specified", "when": "When maximum reasoning effort uses the default adjustment" } ], "pricing": { "billing_mode": "per_request", "image_input_price": null, "image_output_price": null, "per_request_price": null, "unit_prices": { "input_token": 1e-07 } }, "pricing_profiles": [ { "dimensions": { "regime": "base" }, "label": "Base token rates", "units": [ { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_price", "label": "Input tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "output_price", "label": "Output tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "output_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "out", "unit": "output_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_read_price", "label": "Cache read", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "cached_input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "cached_input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_price", "label": "Cache write", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "generic_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "generic_cache_creation_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_1h_price", "label": "Cache write, 1 hour", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "one_hour_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "one_hour_cache_creation_token" } ] }, { "dimensions": { "native_endpoint": "rerank", "quoted_unit": "/ 1K tokens" }, "label": "Ultra", "units": [ { "amount": "0.0001000", "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_token", "label": "input token", "native_amount_usd_per_unit": "0.0000001", "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" } ] } ], "request_model_id": "voyageai/rerank-2.5", "route_label": "Ultra", "status_reads": [], "tiers": [], "units": [ { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_price", "label": "Input tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "output_price", "label": "Output tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "output_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "out", "unit": "output_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_read_price", "label": "Cache read", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "cached_input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "cached_input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_price", "label": "Cache write", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "generic_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "generic_cache_creation_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_1h_price", "label": "Cache write, 1 hour", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "one_hour_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "one_hour_cache_creation_token" }, { "amount": "0.0001000", "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_token", "label": "input token", "native_amount_usd_per_unit": "0.0000001", "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" } ] } } ], "model_id": "voyageai/rerank-2.5", "nativeProvenance": { "input_facts_canonical_sha256": "73a17578c0404f560aa51ff0931203c4638b53ebce175611b90ca13345a3cfeb", "input_html_sha256": "bf51b0a6c2db03a8bfca97be2ea2b5a49d655da6fbe198a07cb3a5fc321b560e", "private_audit_sha256": "e1576c45d1ddc8275865ed7fc44f11da571473b48bad854d5c719a1fb9163394", "role": "historical-normalized-input-audit-hashes", "schema": "ailoot-native-model-provenance-v2" }, "page": { "canonical": "https://ailoot.ai/models/id/voyageai-rerank-2-5/", "description": "rerank-2.5 mengurutkan ulang kumpulan kandidat teks yang diberikan berdasarkan relevansi terhadap kueri dan instruksi bahasa alami opsional. Ini adalah reranker multibahasa yang berorientasi pada kualitas, dengan konteks kueri-dokumen 32.000 token dan skor berperingkat yang terkait dengan indeks kandidat asli.", "locale": "id", "md_url": "https://ailoot.ai/models/id/voyageai-rerank-2-5/index.md", "title": "VoyageAI by MongoDB: rerank-2.5 — Ailoot", "url": "https://ailoot.ai/models/id/voyageai-rerank-2-5/" }, "rawModel": { "about": "rerank-2.5 adalah reranker Voyage AI yang mengikuti instruksi, dirancang untuk menyempurnakan kumpulan kandidat pencarian. Model ini memeriksa kueri dan setiap dokumen secara bersamaan sebagai cross-encoder, lalu mengembalikan urutan relevansi. Rilis pembuat menekankan kendali bahasa alami atas hal yang dianggap relevan, pengambilan informasi multibahasa, dan konteks kueri-dokumen yang lebih panjang daripada edisi rerank-2 sebelumnya.\n\n### Berikan kandidat yang perlu dibandingkan kepada model\n\nMasukannya adalah kueri dan daftar dokumen teks kandidat. Kandidat dapat berupa bagian teks yang dipilih pencarian embedding atau pencarian leksikal tahap pertama. Reranker mengevaluasi teks yang diberikan, bukan membuat embedding korpus, sehingga tugas praktisnya adalah menyempurnakan kandidat yang sudah ditemukan pencarian awal.\n\nPertahankan urutan kandidat asli dan pengidentifikasi sumber saat menyiapkan daftar tersebut. Indeks yang dikembalikan merujuk pada posisi masukan itu. Untuk tugas RAG, berikan bagian teks yang mungkin benar-benar digunakan tahap jawaban, termasuk judul atau kondisi yang diperlukan untuk menafsirkannya. Dokumen relevan yang tidak ada dalam kumpulan kandidat yang dikirim tidak dapat muncul sebagai hasil yang baru ditemukan melalui pemeringkatan ulang.\n\n### Tentukan pengertian relevansi yang dituju\n\nInstruksi dapat ditambahkan sesudah atau sebelum kueri. Rilis tersebut menjelaskan tiga peran yang berguna: menekankan bagian tertentu dari dokumen, mengutamakan jenis sumber, dan memperjelas kueri dengan konteks aplikasi. Ketiganya mengubah tugas relevansi yang disajikan kepada model.\n\nUntuk pencarian teknis, contoh instruksinya adalah “Prioritaskan bagian teks yang menjelaskan perilaku kegagalan dan pemulihan; gambaran pengantar bersifat sekunder.” Untuk kumpulan riset, tentukan apakah judul, metode, atau hasil yang paling penting. Untuk istilah yang ambigu, nyatakan bidang subjek pencarian. Jaga pertanyaan dan instruksi tetap selaras agar yang satu tidak meminta topik luas sementara yang lain diam-diam mengubah tugas.\n\nPreferensi ini memandu pemberian skor relevansi. Jika aplikasi memiliki persyaratan kelayakan ketat—seperti versi sumber, tanggal, atau cakupan akses—siapkan kumpulan kandidat yang memenuhi syarat secara terpisah. Instruksi relevansi model dan pembatasan koleksi deterministik melayani tujuan berbeda.\n\n### Keluaran berperingkat dan penggunaan hasil berikutnya\n\nHasil diurutkan berdasarkan relevansi menurun dan mencakup skor relevansi floating-point serta indeks masukan asli kandidat. Opsi top-k memilih jumlah hasil teratas yang dikembalikan; tanpa opsi tersebut, spesifikasi pembuat mengembalikan hasil untuk semua kandidat yang dikirim. Pengaturan return-documents opsional menyertakan teks sumber yang bersesuaian; jika tidak, respons memuat indeks dan skor.\n\nGunakan urutan tersebut untuk memilih bagian teks yang akan ditampilkan atau digunakan pada tahap jawaban berikutnya, dengan menjaga identitas sumbernya tetap utuh. Periksa teks aktual ketika mengevaluasi peringkat: skor adalah ukuran relevansi terhadap kueri yang dikirim, bukan jawaban yang dihasilkan atau verifikasi bahwa klaim dokumen benar. Ini membuat model berguna untuk menyempurnakan materi pencarian teknis, hukum, keuangan, medis, percakapan, dan multibahasa tanpa mengubah pemeringkatan menjadi kesimpulan bidang tersebut.\n\n### Konteks, anggaran total token, dan truncation\n\nBatas kueri yang didokumentasikan adalah 8.000 token, termasuk instruksi yang ditempatkan dalam kueri. Kueri ditambah satu dokumen kandidat mana pun harus muat dalam 32.000 token, dan daftar dokumen dapat memuat paling banyak 1.000 item. Alokasi total adalah 600.000 token, dihitung sebagai token kueri dikalikan jumlah dokumen, ditambah seluruh token dokumen. Karena itu, instruksi panjang dihitung berulang di seluruh pasangan kandidat.\n\nJumlah top-k keluaran tidak mengubah jumlah kandidat dalam perhitungan token masukan tersebut. Truncation default memendekkan kueri atau dokumen yang terlalu panjang agar memenuhi batas konteks; menonaktifkannya membuat kueri atau pasangan yang terlalu panjang mengembalikan kesalahan. Siapkan bagian teks yang memuat jawaban secara sengaja dan jaga agar kondisi yang relevan tetap berada dalam teks yang akan dilihat model. Evaluasi urutan yang disempurnakan pada pertanyaan dan instruksi representatif, lalu teruskan bagian teks asli yang dipilih ke tahap berikutnya.", "alias_of": null, "aliases": [], "author": "VoyageAI by MongoDB", "available": null, "canonical_model": "voyageai/rerank-2.5", "capabilities": { "context_length": 32000, "manufacturer_input": [ "text" ], "manufacturer_output": [ "rerank" ] }, "card_routes": [ { "label": "Ultra", "rates": [ { "amount": "0.0001000", "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_token", "label": "input token", "native_amount_usd_per_unit": "0.0000001", "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" } ], "summary": "Informasi lengkap harga konfigurasi tersedia di bagian detail." } ], "ctx": null, "entity_type": "model", "formats": [ "rerank" ], "id": "voyageai/rerank-2.5", "input": [ "text" ], "kind": "rerank", "max_out": null, "methods": [ { "available": null, "cache_read": null, "cache_write": null, "cache_write_1h": null, "contract": { "data": { "all_results": "Every actual result index, finite relevance_score and original document object is retained. Indices bind the original document array. Entire original JSON is stored and returned only after settlement.", "errors": [ { "code": null, "condition": "Bidang wajib, representasi, atau nilai tidak valid", "http_status": 400, "recovery": "Perbaiki bidang permintaan yang ditunjukkan.", "type": "invalid_request_error" } ], "hints": [], "ignored_fields": [], "lifecycle": { "cancel": null, "idempotency_header": null, "job_retrieve": null, "mode": "synchronous", "streaming": false, "unknown": "Simpan operasi awal dan balasan lengkap/sebagian; jangan ulangi POST." }, "material_combinations": [ { "accepted": [ "string documents", "objects with text; image only for separately exact-supported multimodal model" ], "mode": "text_rerank", "required": [ "query", "documents array with minItems1" ] } ], "operation_modes": [ "rerank" ], "outputs": { "all_results": "Setiap indeks hasil aktual, relevance_score yang bernilai berhingga, dan objek dokumen asli dipertahankan. Indeks mengacu pada array dokumen asli. Seluruh JSON asli disimpan dan hanya dikembalikan setelah penyelesaian tagihan.", "family": "rerank", "fields": [ { "description": null, "name": "model", "type": "string" }, { "description": null, "name": "id", "type": "string" }, { "description": null, "name": "provider", "type": "string" }, { "description": null, "name": "results[].index", "type": "integer" }, { "description": null, "name": "results[].relevance_score", "type": "number" }, { "description": null, "name": "results[].document.text", "type": "string" }, { "description": null, "name": "results[].document.image", "type": "string" }, { "description": null, "name": "usage.search_units", "type": "integer" }, { "description": null, "name": "usage.total_tokens", "type": "integer" }, { "description": null, "name": "usage.cost", "type": "number" } ], "format": "json", "schema": { "description": "Rerank response containing ranked results", "example": { "id": "gen-rerank-1234567890-abc", "model": "cohere/rerank-v3.5", "results": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "usage": { "search_units": 1, "total_tokens": 150 } }, "properties": { "id": { "description": "Unique identifier for the rerank response (ORID format)", "example": "gen-rerank-1234567890-abc", "type": "string" }, "model": { "description": "The model used for reranking", "example": "cohere/rerank-v3.5", "type": "string" }, "provider": { "description": "The provider that served the rerank request", "example": "Cohere", "type": "string" }, "results": { "description": "List of rerank results sorted by relevance", "example": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "items": { "description": "A single rerank result", "example": { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 }, "properties": { "document": { "description": "The document object echoing the original input (text and/or image)", "properties": { "image": { "description": "The image (URL or data URI) from the original document", "example": "https://example.com/image.png", "type": "string" }, "text": { "description": "The document text", "example": "Paris is the capital of France.", "type": "string" } }, "type": "object" }, "index": { "description": "Index of the document in the original input list", "example": 0, "type": "integer" }, "relevance_score": { "description": "Relevance score of the document to the query", "example": 0.98, "format": "double", "type": "number" } }, "required": [ "index", "relevance_score", "document" ], "type": "object" }, "type": "array" }, "usage": { "description": "Usage statistics", "example": { "search_units": 1, "total_tokens": 150 }, "properties": { "cost": { "description": "Cost of the request in credits", "example": 0.001, "format": "double", "type": "number" }, "search_units": { "description": "Number of search units consumed (Cohere billing)", "example": 1, "type": "integer" }, "total_tokens": { "description": "Total number of tokens used", "example": 150, "type": "integer" } }, "type": "object" } }, "required": [ "model", "results" ], "type": "object" } }, "parameters": [ { "availability": "source_interface", "default": null, "default_known": false, "description": "The list of documents to rerank. Documents may be plain strings, or structured objects with `text` and/or `image` for multimodal models.", "enum": null, "enum_known": false, "minItems": 1, "name": "documents", "required": true, "type": "array" }, { "availability": "source_interface", "default": null, "default_known": false, "enum": [ "voyageai/rerank-2.5" ], "enum_known": true, "name": "model", "required": true, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "enum": null, "enum_known": false, "name": "provider", "required": false, "type": "schema" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "The search query to rerank documents against", "enum": null, "enum_known": false, "name": "query", "required": true, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "A unique identifier for grouping related requests (e.g., a conversation or agent workflow). Used for observability grouping in Broadcast and private logging; never sent to the provider. If provided in both the request body and the x-session-id header, the body value takes precedence. Maximum of 256 characters.", "enum": null, "enum_known": false, "maxLength": 256, "name": "session_id", "required": false, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "Number of most relevant documents to return", "enum": null, "enum_known": false, "minimum": 1, "name": "top_n", "required": false, "type": "integer" }, { "availability": "source_interface", "default": null, "default_known": false, "enum": null, "enum_known": false, "name": "trace", "required": false, "type": "schema" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "A unique identifier representing your end-user. Forwarded to Broadcast and private logging as the end-user id; never sent to the provider.", "enum": null, "enum_known": false, "maxLength": 256, "name": "user", "required": false, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "enum": null, "enum_known": false, "name": "documents[].text", "required": false, "type": "string" }, { "availability": "source_interface", "default": null, "default_known": false, "description": "Official family image carrier is remote http/https URL or data:image URI; only for exact multimodal-model bindings.", "enum": null, "enum_known": false, "name": "documents[].image", "required": false, "type": "string" } ], "unsupported": [ "stream=true" ] }, "hints": [ "Kirim ID model yang persis dan simpan respons native secara utuh." ], "input": { "content_type": "application/json", "material_types": [ "text" ], "multimodal_input_binding": null, "optional_controls": [ "provider", "session_id", "top_n", "trace", "user" ], "required": [ "model", "query", "documents" ] }, "lifecycle": { "cancel": null, "idempotency_header": null, "job_retrieve": null, "mode": "synchronous", "streaming": false, "unknown": "Preserve original operation and full/partial reply; do not replay POST." }, "manufacturer_capability": { "context_length": 32000, "manufacturer_input": [ "text" ], "manufacturer_output": [ "rerank" ] }, "method": { "http": "POST", "path": "/v1/rerank" }, "method_aliases": [], "operation_modes": [ "rerank" ], "output": { "all_results": "Every actual result index, finite relevance_score and original document object is retained. Indices bind the original document array. Entire original JSON is stored and returned only after settlement.", "family": "rerank", "fields": [ { "description": null, "name": "model", "type": "string" }, { "description": null, "name": "id", "type": "string" }, { "description": null, "name": "provider", "type": "string" }, { "description": null, "name": "results[].index", "type": "integer" }, { "description": null, "name": "results[].relevance_score", "type": "number" }, { "description": null, "name": "results[].document.text", "type": "string" }, { "description": null, "name": "results[].document.image", "type": "string" }, { "description": null, "name": "usage.search_units", "type": "integer" }, { "description": null, "name": "usage.total_tokens", "type": "integer" }, { "description": null, "name": "usage.cost", "type": "number" } ], "format": "json", "schema": { "description": "Rerank response containing ranked results", "example": { "id": "gen-rerank-1234567890-abc", "model": "cohere/rerank-v3.5", "results": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "usage": { "search_units": 1, "total_tokens": 150 } }, "properties": { "id": { "description": "Unique identifier for the rerank response (ORID format)", "example": "gen-rerank-1234567890-abc", "type": "string" }, "model": { "description": "The model used for reranking", "example": "cohere/rerank-v3.5", "type": "string" }, "provider": { "description": "The provider that served the rerank request", "example": "Cohere", "type": "string" }, "results": { "description": "List of rerank results sorted by relevance", "example": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "items": { "description": "A single rerank result", "example": { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 }, "properties": { "document": { "description": "The document object echoing the original input (text and/or image)", "properties": { "image": { "description": "The image (URL or data URI) from the original document", "example": "https://example.com/image.png", "type": "string" }, "text": { "description": "The document text", "example": "Paris is the capital of France.", "type": "string" } }, "type": "object" }, "index": { "description": "Index of the document in the original input list", "example": 0, "type": "integer" }, "relevance_score": { "description": "Relevance score of the document to the query", "example": 0.98, "format": "double", "type": "number" } }, "required": [ "index", "relevance_score", "document" ], "type": "object" }, "type": "array" }, "usage": { "description": "Usage statistics", "example": { "search_units": 1, "total_tokens": 150 }, "properties": { "cost": { "description": "Cost of the request in credits", "example": 0.001, "format": "double", "type": "number" }, "search_units": { "description": "Number of search units consumed (Cohere billing)", "example": 1, "type": "integer" }, "total_tokens": { "description": "Total number of tokens used", "example": 150, "type": "integer" } }, "type": "object" } }, "required": [ "model", "results" ], "type": "object" } }, "schema": { "request": { "description": "Rerank request input", "example": { "documents": [ "Paris is the capital of France.", "Berlin is the capital of Germany." ], "model": "cohere/rerank-v3.5", "query": "What is the capital of France?", "top_n": 3 }, "properties": { "documents": { "description": "The list of documents to rerank. Documents may be plain strings, or structured objects with `text` and/or `image` for multimodal models.", "example": [ "Paris is the capital of France.", "Berlin is the capital of Germany." ], "items": { "anyOf": [ { "type": "string" }, { "description": "A structured document with optional text and/or image content. At least one of `text` or `image` must be provided.", "properties": { "image": { "description": "An image associated with the document, as a remote URL (http/https) or a base64-encoded data URI (data:image/...).", "example": "https://upload.wikimedia.org/wikipedia/commons/thumb/8/8b/Phytogenic.png", "type": "string" }, "text": { "description": "The document text", "example": "AI enables robots to perceive, plan, and act autonomously.", "type": "string" } }, "type": "object" } ], "description": "A document to rerank. Either a plain string, or a structured object with optional `text` and/or `image`." }, "minItems": 1, "type": "array" }, "model": { "enum": [ "voyageai/rerank-2.5" ], "type": "string" }, "provider": { "allOf": [ { "$ref": "#/components/schemas/ProviderPreferences" }, { "description": "Provider routing preferences for the request." } ] }, "query": { "description": "The search query to rerank documents against", "example": "What is the capital of France?", "type": "string" }, "session_id": { "description": "A unique identifier for grouping related requests (e.g., a conversation or agent workflow). Used for observability grouping in Broadcast and private logging; never sent to the provider. If provided in both the request body and the x-session-id header, the body value takes precedence. Maximum of 256 characters.", "example": "session-1234", "maxLength": 256, "type": "string" }, "top_n": { "description": "Number of most relevant documents to return", "example": 3, "minimum": 1, "type": "integer" }, "trace": { "$ref": "#/components/schemas/TraceConfig" }, "user": { "description": "A unique identifier representing your end-user. Forwarded to Broadcast and private logging as the end-user id; never sent to the provider.", "example": "user-1234", "maxLength": 256, "type": "string" } }, "required": [ "model", "query", "documents" ], "type": "object" }, "response": { "description": "Rerank response containing ranked results", "example": { "id": "gen-rerank-1234567890-abc", "model": "cohere/rerank-v3.5", "results": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "usage": { "search_units": 1, "total_tokens": 150 } }, "properties": { "id": { "description": "Unique identifier for the rerank response (ORID format)", "example": "gen-rerank-1234567890-abc", "type": "string" }, "model": { "description": "The model used for reranking", "example": "cohere/rerank-v3.5", "type": "string" }, "provider": { "description": "The provider that served the rerank request", "example": "Cohere", "type": "string" }, "results": { "description": "List of rerank results sorted by relevance", "example": [ { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 } ], "items": { "description": "A single rerank result", "example": { "document": { "text": "Paris is the capital of France." }, "index": 0, "relevance_score": 0.98 }, "properties": { "document": { "description": "The document object echoing the original input (text and/or image)", "properties": { "image": { "description": "The image (URL or data URI) from the original document", "example": "https://example.com/image.png", "type": "string" }, "text": { "description": "The document text", "example": "Paris is the capital of France.", "type": "string" } }, "type": "object" }, "index": { "description": "Index of the document in the original input list", "example": 0, "type": "integer" }, "relevance_score": { "description": "Relevance score of the document to the query", "example": 0.98, "format": "double", "type": "number" } }, "required": [ "index", "relevance_score", "document" ], "type": "object" }, "type": "array" }, "usage": { "description": "Usage statistics", "example": { "search_units": 1, "total_tokens": 150 }, "properties": { "cost": { "description": "Cost of the request in credits", "example": 0.001, "format": "double", "type": "number" }, "search_units": { "description": "Number of search units consumed (Cohere billing)", "example": 1, "type": "integer" }, "total_tokens": { "description": "Total number of tokens used", "example": 150, "type": "integer" } }, "type": "object" } }, "required": [ "model", "results" ], "type": "object" } } }, "ctx": null, "declared_pricing_profiles": [ { "dimensions": { "regime": "base" }, "label": "Base token rates", "units": [ { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_price", "label": "Input tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "output_price", "label": "Output tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "output_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "out", "unit": "output_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_read_price", "label": "Cache read", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "cached_input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "cached_input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_price", "label": "Cache write", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "generic_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "generic_cache_creation_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_1h_price", "label": "Cache write, 1 hour", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "one_hour_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "one_hour_cache_creation_token" } ] }, { "dimensions": { "native_endpoint": "rerank", "quoted_unit": "/ 1K tokens" }, "label": "Ultra", "units": [ { "amount": "0.0001000", "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_token", "label": "input token", "native_amount_usd_per_unit": "0.0000001", "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" } ] } ], "examples": [ { "code": "{\n \"documents\": [\n \"Paris is the capital of France.\",\n \"Berlin is the capital of Germany.\"\n ],\n \"model\": \"voyageai/rerank-2.5\",\n \"query\": \"What is the capital of France?\",\n \"top_n\": 3\n}", "label": "Retained official request structure; documentary example", "language": "json" }, { "code": "set -eu\n: \"${AILOOT_API_KEY:?Set your Ailoot API key}\"\n: \"${AILOOT_REQUEST_FILE:?Set your actual JSON request file}\"\nexport AILOOT_REQUEST_FILE\npython3 - <<'PY_REQUEST'\nimport json,os\nfrom pathlib import Path\np=json.loads(Path(os.environ['AILOOT_REQUEST_FILE']).read_text(encoding='utf-8'))\nif not isinstance(p,dict): raise SystemExit('Request must be a JSON object')\np['model']='voyageai/rerank-2.5'\nPath('request.json').write_text(json.dumps(p,ensure_ascii=False),encoding='utf-8')\nPY_REQUEST\ncurl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/rerank \\\n -H \"Authorization: Bearer $AILOOT_API_KEY\" -H 'Content-Type: application/json' \\\n --data-binary @request.json -D response-headers.txt -o result.json\npython3 - <<'PY_RESULT'\nimport json\nfrom pathlib import Path\nprint(json.dumps(json.loads(Path('result.json').read_text()),ensure_ascii=False,indent=2))\nPY_RESULT\n", "label": "Use your actual request material and retain the complete result", "language": "bash", "material_policy": "Use real owner-supplied state/questions or query/documents. No invented file ID, document, image URL, job ID or returned result." } ], "family": "rerank", "fast_multiplier": null, "flex_multiplier": null, "fmt": "rerank", "http": "POST", "in": null, "input": [ "text" ], "label": "Pemeringkatan ulang", "max_out": null, "max_reasoning_effort_multiplier": null, "method_aliases": [], "native_endpoint": "rerank", "out": null, "output": [ "rerank" ], "path": "/v1/rerank", "price_conditions": [ { "label": "Fast mode", "value": "Not specified", "when": "When fast mode is selected" }, { "label": "Flex mode", "value": "Not specified", "when": "When flex mode is selected" }, { "label": "Maximum reasoning effort", "value": "Not specified", "when": "When maximum reasoning effort uses the default adjustment" } ], "pricing": { "billing_mode": "per_request", "image_input_price": null, "image_output_price": null, "per_request_price": null, "unit_prices": { "input_token": 1e-07 } }, "pricing_profiles": [ { "dimensions": { "regime": "base" }, "label": "Base token rates", "units": [ { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_price", "label": "Input tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "output_price", "label": "Output tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "output_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "out", "unit": "output_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_read_price", "label": "Cache read", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "cached_input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "cached_input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_price", "label": "Cache write", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "generic_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "generic_cache_creation_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_1h_price", "label": "Cache write, 1 hour", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "one_hour_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "one_hour_cache_creation_token" } ] }, { "dimensions": { "native_endpoint": "rerank", "quoted_unit": "/ 1K tokens" }, "label": "Ultra", "units": [ { "amount": "0.0001000", "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_token", "label": "input token", "native_amount_usd_per_unit": "0.0000001", "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" } ] } ], "request_model_id": "voyageai/rerank-2.5", "route_label": "Ultra", "status_reads": [], "tiers": [], "units": [ { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_price", "label": "Input tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "output_price", "label": "Output tokens", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "output_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "out", "unit": "output_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_read_price", "label": "Cache read", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "cached_input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "cached_input_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_price", "label": "Cache write", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "generic_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "generic_cache_creation_token" }, { "amount": null, "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "cache_write_1h_price", "label": "Cache write, 1 hour", "native_amount_usd_per_unit": null, "native_quantity": 1, "native_unit": "one_hour_cache_creation_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "one_hour_cache_creation_token" }, { "amount": "0.0001000", "amount_unit": "usd_per_thousand_tokens", "billing_input": true, "estimate_only": false, "key": "input_token", "label": "input token", "native_amount_usd_per_unit": "0.0000001", "native_quantity": 1, "native_unit": "input_token", "public_unit_label": "/ 1K tokens", "quantity": 1000, "quantity_is_estimate": false, "rate_role": "typed_native_retail_rate", "side": "in", "unit": "input_token" } ] } ], "model_aliases": [ { "canonical_model_id": "voyageai/rerank-2.5", "request_model_id": "voyageai/rerank-2.5", "route_label": "Ultra" } ], "name": "VoyageAI by MongoDB: rerank-2.5", "price": { "in": null, "in_label": null, "out": null, "out_label": null, "sort_in": null, "sort_out": null, "varies": true }, "public_id": "voyageai/rerank-2.5", "short": "rerank-2.5 mengurutkan ulang kumpulan kandidat teks yang diberikan berdasarkan relevansi terhadap kueri dan instruksi bahasa alami opsional. Ini adalah reranker multibahasa yang berorientasi pada kualitas, dengan konteks kueri-dokumen 32.000 token dan skor berperingkat yang terkait dengan indeks kandidat asli." }, "renderer_schema": { "method_order": "Original DTO methods order; every exact route/format/path/caller survives.", "prices": "Original public DTO fields and original native display HTML; no flattened or normalized offers." }, "representation": "exact-public-dto-and-native-renderer-html", "representationNotes": { "current_reference": { "about_json_pointer": "/aboutHtml", "methods": [ { "contract_json_pointer": "/methods/0/rawMethod/contract", "html_json_pointer": "/methods/0/html", "method_ordinal": 0 } ], "role": "current-semantic-reference" }, "raw_input": { "examples_role": "historical-input-not-current-ready-examples", "json_pointers": [ "/rawModel", "/methods/0/rawMethod" ], "role": "historical-owner-input" }, "schema": "ailoot-native-model-representations-v1" }, "schema": "ailoot-static-model-facts-v1", "schema_version": 1, "shortText": "rerank-2.5 mengurutkan ulang kumpulan kandidat teks yang diberikan berdasarkan relevansi terhadap kueri dan instruksi bahasa alami opsional. Ini adalah reranker multibahasa yang berorientasi pada kualitas, dengan konteks kueri-dokumen 32.000 token dan skor berperingkat yang terkait dengan indeks kandidat asli.", "source": { "audit_sha256": "2e00086bcbfc28e64d0cebe97aa47c17d8d5618963027f8ffb9dc6123d7d25b1", "bridge_map_sha256": "1125275768b9636fc8461bc7da5061239a81ddfe508e2557843ca42c58ab0036", "raw_model_json_sha256": "497c79b252fc940fdb8032803369c1a6aadaa1636f3c572ba6064e18b6601ced", "renderer_sha256": "2a0471475a29000c23db326ca1e315df73dee7524f6bc2907818ee3458813f2d" }, "sourceRevision": "089f0afb15354486f855578d5b46f48eab7933e9d30d02355962722559e64bb3" }