NVIDIA: Nemotron 3.5 Super VL

{"about":"### Penalaran, kode, dan teks panjang\n\nNemotron 3.5 Super VL adalah model NVIDIA untuk tugas yang menghubungkan beberapa langkah penalaran, memahami kode program, atau mengolah materi besar. Di Ailoot API, kartu ini menyediakan dialog teks: Anda mengirim instruksi, data awal, dan riwayat pesan, lalu model mengembalikan jawaban teks.\n\nModel dapat digunakan untuk menelaah dokumentasi teknis, membandingkan persyaratan, menjelaskan kode, dan menganalisis laporan teks. Tentukan tujuan, lampirkan materi yang diperlukan, serta bentuk hasil, misalnya daftar kesimpulan beralasan, rencana perubahan program, atau perbandingan beberapa solusi.\n\n### Arsitektur dan spesialisasi\n\nSuper VL asal adalah model visi-bahasa NVIDIA dengan encoder visual C-RADIO yang juga dirancang untuk memahami dokumen dan video. Kartu Ailoot API ini menyediakan pekerjaan teksnya; materi awal dikirim sebagai isi teks.\n\nDasar model menggabungkan Mamba-2, Transformer, dan arsitektur Mixture of Experts. Dari sekitar 120 miliar parameter, sekitar 12 miliar aktif pada setiap langkah pemrosesan. Struktur model ini perlu dipertimbangkan terpisah dari kecepatan aktual permintaan tertentu.\n\nPelatihan menggunakan repositori perangkat lunak, pasangan tugas dan perubahan kode, log build, lintasan pemecahan tugas bertahap, serta materi konteks panjang. Pelatihan lanjutan menggabungkan kode, penalaran, instruksi, dan dialog. Spesialisasi ini berguna untuk tugas gabungan yang perlu menghubungkan informasi awal sebelum menyusun hasil.\n\n### Menyiapkan tugas\n\nUntuk analisis dokumen, rumuskan pertanyaan dan kriteria pemilihan informasi terlebih dahulu, lalu lampirkan teks dengan judul bagian. Jika membandingkan sumber, beri label singkat dan minta kesimpulan dikaitkan dengan label tersebut. Untuk kode, tambahkan potongan program terkait, perilaku teramati, dan hasil yang diharapkan agar jawaban dapat diperiksa pada kasus tertentu.\n\nJendela konteks besar membantu menempatkan materi terkait dalam satu permintaan. Volume konteks tidak menggantikan tugas yang jelas: pisahkan instruksi dari data awal dan sebutkan solusi atau dokumen yang diinginkan sejak awal.\n\n### Konteks dan cara koneksi\n\nDi Ailoot API, varian ini memiliki jendela konteks 262144 token. Riwayat percakapan, instruksi, materi awal, dan jawaban berbagi ruang tersebut; sisakan ruang hasil saat menyiapkan permintaan panjang.\n\nModel tersedia melalui Chat Completions, Responses, Messages, dan antarmuka teks berformat Gemini. Alur varian ini adalah masukan teks dan keluaran teks; dokumen dikirim sebagai isi teks.","alias_of":null,"aliases":[],"author":"NVIDIA","canonical_model":"nvidia/nemotron-3.5-super-vl","ctx":null,"formats":["chat","gemini","messages","responses"],"id":"nvidia/nemotron-3.5-super-vl","input":["text"],"kind":"text","max_out":null,"methods":[{"cache_read":0.04,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":null,"family":"anthropic","fast_multiplier":1,"flex_multiplier":1,"fmt":"messages","http":"POST","img_in":null,"img_out":null,"in":0.3,"input":["text"],"label":"Anthropic Messages","max_out":null,"max_tool_calls":null,"out":1.0,"output":["text"],"path":"/v1/messages","request_model_id":"nvidia/nemotron-3.5-super-vl","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]},{"cache_read":0.04,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":null,"family":"openai","fast_multiplier":1,"flex_multiplier":1,"fmt":"responses","http":"POST","img_in":null,"img_out":null,"in":0.3,"input":["text"],"label":"OpenAI Responses","max_out":null,"max_tool_calls":null,"out":1.0,"output":["text"],"path":"/v1/responses","request_model_id":"nvidia/nemotron-3.5-super-vl","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]},{"cache_read":0.04,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":null,"family":"openai","fast_multiplier":1,"flex_multiplier":1,"fmt":"chat","http":"POST","img_in":null,"img_out":null,"in":0.3,"input":["text"],"label":"OpenAI Chat","max_out":null,"max_tool_calls":null,"out":1.0,"output":["text"],"path":"/v1/chat/completions","request_model_id":"nvidia/nemotron-3.5-super-vl","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]},{"cache_read":0.04,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":null,"family":"gemini","fast_multiplier":1,"flex_multiplier":1,"fmt":"gemini","http":"POST","img_in":null,"img_out":null,"in":0.3,"input":["text"],"label":"Gemini generateContent","max_out":null,"max_tool_calls":null,"out":1.0,"output":["text"],"path":"/v1beta/models/nvidia/nemotron-3.5-super-vl:generateContent","request_model_id":"nvidia/nemotron-3.5-super-vl","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]}],"model_aliases":[{"canonical_model_id":"nvidia/nemotron-3.5-super-vl","request_model_id":"nvidia/nemotron-3.5-super-vl","route_label":"Ultra"}],"name":"NVIDIA: Nemotron 3.5 Super VL","price":{"in":0.3,"in_label":null,"out":1.0,"out_label":null,"sort_in":0.3,"sort_out":1.0,"varies":false},"public_id":"nvidia/nemotron-3.5-super-vl","short":"Dialog teks dengan model NVIDIA untuk penalaran, pekerjaan kode, dan analisis materi besar."}