Source: https://ailoot.ai/models/id/nvidia-llama-3-1-nemotron-safety-guard-8b-v3/
Canonical: https://ailoot.ai/models/id/nvidia-llama-3-1-nemotron-safety-guard-8b-v3/
Language: id

# Llama 3.1 Nemotron Safety Guard 8B v3

[Markdown](https://ailoot.ai/models/id/nvidia-llama-3-1-nemotron-safety-guard-8b-v3/index.md) · [JSON](https://ailoot.ai/models/id/nvidia-llama-3-1-nemotron-safety-guard-8b-v3/facts.json)

## [Llama 3.1 Nemotron Safety Guard 8B v3](https://ailoot.ai/models/id/nvidia-llama-3-1-nemotron-safety-guard-8b-v3/)

Nemotron Safety Guard v3 memeriksa pesan pengguna dan jawaban model berdasarkan taksonomi risiko yang ditentukan. Mengembalikan JSON dengan penilaian keselamatan terpisah dan kategori pelanggaran.

- [**Ailoot · OpenAI Chat** `POST /v1/chat/completions` `nvidia/llama-3.1-nemotron-safety-guard-8b-v3`](https://ailoot.ai/models/id/nvidia-llama-3-1-nemotron-safety-guard-8b-v3/#method-0)
- [**Ailoot · OpenAI Chat** `POST /v1/chat/completions` `nemotron-safety-guard`](https://ailoot.ai/models/id/nvidia-llama-3-1-nemotron-safety-guard-8b-v3/#method-1)
- [**Ailoot · OpenAI Chat** `POST /v1/chat/completions` `nemotron-safety-guard-latest`](https://ailoot.ai/models/id/nvidia-llama-3-1-nemotron-safety-guard-8b-v3/#method-2)

Tentang model

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

#### Memeriksa kedua sisi percakapan

Llama 3.1 Nemotron Safety Guard 8B v3 dikembangkan NVIDIA untuk memoderasi interaksi manusia dengan model bahasa. Masukan dapat berupa pesan pengguna saja atau dilengkapi jawaban asisten, termasuk beberapa giliran. Model menentukan keselamatan masing-masing sisi: permintaan berbahaya dan penolakan asisten yang tepat mendapat penilaian berbeda. Ini memungkinkan pemeriksaan masukan sebelum generasi dan analisis terpisah atas jawaban yang telah disiapkan.

Hasil berupa JSON. Bidang wajib `User Safety` memuat `safe` atau `unsafe`. Bidang `Response Safety` muncul ketika masukan berisi jawaban asisten. Jika pelanggaran ditemukan, `Safety Categories` ditambahkan dengan daftar kategori yang dipisahkan koma. Hasil tersebut sesuai untuk penyaringan program dan penerusan kasus yang diperdebatkan ke pemeriksaan tambahan.

#### Kategori dan konteks budaya

Instruksi menetapkan aturan serta nama kategori yang digunakan untuk pemeriksaan. Taksonomi yang dipublikasikan memiliki 23 kelompok risiko, termasuk kekerasan, ancaman, perundungan, pengungkapan data pribadi, penipuan, dan kode berbahaya. Nama dalam hasil bergantung pada kebijakan yang diberikan, sehingga perlu ditafsirkan bersama aturan tugas tertentu.

Pelatihan mempertimbangkan karakteristik regional topik sensitif. Data disesuaikan dengan ungkapan setempat dan konteks budaya, sedangkan terjemahan yang tidak konsisten dibuang. Pelatihan mencakup permintaan provokatif dan upaya mendorong asisten melanggar aturan. Susunan pelatihan tersebut sangat berguna ketika terjemahan harfiah kata terlarang tidak cukup untuk menilai makna pesan.

#### Bahasa dan jumlah materi

Sembilan bahasa utama adalah Inggris, Spanyol, Tionghoa, Jerman, Prancis, Hindi, Jepang, Arab, dan Thai. NVIDIA juga menjelaskan generalisasi ke bahasa lain, termasuk Rusia; ini merupakan sifat terpisah, bukan penyertaan Rusia dalam sembilan bahasa utama. Instruksi umum dipertahankan antarbahasa, sedangkan giliran yang diperiksa diberikan dalam bahasa percakapan.

Masukan berupa teks, dengan konteks hingga 8K token. Fondasi model adalah Llama 3.1 8B Instruct yang disetel lebih lanjut untuk tugas keselamatan. Keluaran berfungsi sebagai klasifikasi berdasarkan aturan yang ditentukan: keluaran perlu diuji pada pesan yang khas bagi aplikasi dan tidak menjamin moderasi tanpa kesalahan.

Tugas praktis

### Tugas praktis

Contoh teknis dalam bahasa Inggris

```
Moderate a single user message

Assess this single user message for safety: 'Please help me write a thank-you note to the volunteers who cleaned our local park.' Return the documented safety assessment for user input alone using the User Safety field. No assistant response is supplied; use the documented safe or unsafe values.
```

#### Hasil yang diharapkan

```
A JSON safety assessment using the documented User Safety field for the supplied user message.
```

#### Periksa hasil

```
User Safety must use safe or unsafe; do not invent a custom classification label or assess an absent assistant response.
```

**Ailoot · OpenAI Chat** `POST /v1/chat/completions` `nvidia/llama-3.1-nemotron-safety-guard-8b-v3`

[`nvidia/llama-3.1-nemotron-safety-guard-8b-v3`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&fmt=chat)

NVIDIA /`nvidia/llama-3.1-nemotron-safety-guard-8b-v3`

ID model

- `nvidia/llama-3.1-nemotron-safety-guard-8b-v3`
- `nemotron-safety-guard`
- `nemotron-safety-guard-latest`

### Mulai

Nemotron Safety Guard v3 memeriksa pesan pengguna dan jawaban model berdasarkan taksonomi risiko yang ditentukan. Mengembalikan JSON dengan penilaian keselamatan terpisah dan kategori pelanggaran.

Masukan yang diterima

Teks

Hasil

Teks

### Tersedia di Ailoot · masukan & metode

**Ailoot · OpenAI Chat** `POST /v1/chat/completions`

Masukan

`text`

Keluaran

`text`

Konteks

4.096 tokens

Keluaran maksimum

1.024 tokens

Format keluaran

json atau SSE

| Kolom | Jenis | Persyaratan | Nilai & batasan | Bawaan |
| --- | --- | --- | --- | --- |
| Kolom`model` | Jenis`string` | PersyaratanWajib | Nilai & batasan nvidia/llama-3.1-nemotron-safety-guard-8b-v3 | Bawaan— |
| Kolom`messages` | Jenis`message[]` | PersyaratanWajib | Nilai & batasan Item minimum: 1 | Bawaan— |
| Kolom`messages[].role` | Jenis`string` | PersyaratanWajib | Nilai & batasan Contoh: system, user, assistant, tool, function Kondisi: Peran ini direpresentasikan secara eksplisit oleh tipe kompatibilitas dalam kode sumber; kumpulan peran model yang tepat tetap khusus untuk tiap model. Bergantung pada model | Bawaan— |
| Kolom`messages[].content` | Jenis`string\|content_part[]\|null` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`messages[].tool_calls` | Jenis`tool_call[]` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`messages[].tool_call_id` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Berlaku jika: role=tool; gunakan ID yang benar-benar dikembalikan | Bawaan— |
| Kolom`messages[].name` | Jenis`string` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`messages[].reasoning_content` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Berlaku jika: Kirim kembali penalaran yang benar-benar dikembalikan jika model ini menggunakannya | Bawaan— |
| Kolom`messages[].reasoning_details` | Jenis`object\|array` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`max_tokens` | Jenis`integer` | PersyaratanOpsional | Nilai & batasan Minimum: 1 Bergantung pada model | Bawaan— |
| Kolom`max_completion_tokens` | Jenis`integer` | PersyaratanOpsional | Nilai & batasan Minimum: 1 Bergantung pada model | Bawaan— |
| Kolom`stream` | Jenis`boolean` | PersyaratanOpsional | Nilai & batasan | Bawaanfalse |
| Kolom`stream_options.include_usage` | Jenis`boolean` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`temperature` | Jenis`number` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`top_p` | Jenis`number` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`stop` | Jenis`string\|string[]` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`tools` | Jenis`function_tool[]` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`tools[].type` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`tools[].function.name` | Jenis`string` | PersyaratanWajib | Nilai & batasan Berlaku jika: alat fungsi disediakan | Bawaan— |
| Kolom`tools[].function.description` | Jenis`string` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`tools[].function.parameters` | Jenis`JSON_schema` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`tool_choice` | Jenis`string\|object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`parallel_tool_calls` | Jenis`boolean` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`response_format` | Jenis`object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`reasoning_effort` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`reasoning` | Jenis`object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`service_tier` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`instructions` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Kolom kompatibilitas | Bawaan— |
| Kolom`functions` | Jenis`function[]` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`function_call` | Jenis`string\|object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`prompt_cache_options` | Jenis`object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |

Kombinasi bahan masukan

**Teks** · Teks UTF8

`messages[].content`

Parameter metode dan siklus hidup

| Parameter | Kontrak |
| --- | --- |
| Parameter`material_types` | Kontraktext |
| Parameter`required` | Kontrakmodel, messages |
| Parameter`optional_controls` | Kontraknull |
| Parameter`content_type` | Kontrakapplication/json |
| Parameter`text_file_material` | Kontrakkonten UTF-8 aktual dimasukkan sebagai teks; ID file atau alamat URL (URLs) memerlukan dukungan yang tepat dan terpisah |

Siklus hidup: `exact retained published method`

Struktur metode dan ID bersifat eksak; kemampuan pemasok terpisah dari penerimaan permintaan.

### Paket

#### Tarif metode Ailoot

Masukan

US$0,0002 */ 1K*

Keluaran

US$0,0002 */ 1K*

Pembacaan cache

US$0,0002 */ 1K*

Penulisan cache

US$0,0002 */ 1K*

Penulisan cache, 5 menit

US$0,0002 */ 1K tokens*

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 →](https://ailoot.ai/docs/id/billing/reserve/)

### Contoh permintaan

#### Ailoot · OpenAI Chat

Contoh teksPython SDKNode.js SDKAliran Python

```bash
set -eu
: "${AILOOT_API_KEY:?Set your Ailoot API key}"
: "${AILOOT_TEXT_FILE:?Set your actual UTF-8 source-material file}"
export AILOOT_TEXT_FILE
python3 - <<'PY_REQUEST'
import json,os
from pathlib import Path
p=json.loads('{"model": "nvidia/llama-3.1-nemotron-safety-guard-8b-v3", "messages": []}')
f=Path(os.environ["AILOOT_TEXT_FILE"])
if not f.is_file():raise SystemExit("Choose your actual UTF-8 source material")
text=f.read_text(encoding="utf-8")
if not text.strip():raise SystemExit("Source material must contain text")
p["messages"]=[{"role":"user","content":"Summarize this material and quote supporting source passages.\n\n"+text}]
Path("request.json").write_text(json.dumps(p,ensure_ascii=False))
PY_REQUEST
curl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/chat/completions \
  -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \
  --data-binary @request.json -D response-headers.txt -o result.json
python3 - <<'PY_RESULT'
import json
from pathlib import Path
print(json.dumps(json.loads(Path("result.json").read_text()),ensure_ascii=False,indent=2))
PY_RESULT
```

```python
from openai import OpenAI
import os

if not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")
client = OpenAI(base_url="https://api.ailoot.ai/v1", api_key=os.environ["AILOOT_API_KEY"], max_retries=0)
resp = client.chat.completions.create(
    model="nvidia/llama-3.1-nemotron-safety-guard-8b-v3",
    messages=[{"role": "user", "content": "Explain the difference between an observation and a hypothesis in three sentences."}],
)
import base64, json
from pathlib import Path

def save_result(data, stem="result"):
    # Keep every field, including usage, IDs, reasoning and tool requests.
    Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
    print(json.dumps(data, ensure_ascii=False, indent=2))
    def save_blob(encoded, suffix, label):
        raw = base64.b64decode(encoded, altchars=b"-_", validate=True)
        ext = "png" if raw.startswith(b"\x89PNG\r\n\x1a\n") else "jpg" if raw.startswith(b"\xff\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix
        file = Path(stem + "-" + label + "." + ext)
        file.write_bytes(raw)
        print("Saved", file)
    for index, block in enumerate(data.get("content") or []):
        print("Content", index, json.dumps(block, ensure_ascii=False))
    for choice in data.get("choices") or []:
        print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))
    for index, item in enumerate(data.get("output") or []):
        print("Output", index, json.dumps(item, ensure_ascii=False))
        if item.get("type") == "image_generation_call" and item.get("result"):
            save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))
    for ci, candidate in enumerate(data.get("candidates") or []):
        for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):
            print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))
            blob = part.get("inlineData") or part.get("inline_data")
            if blob and blob.get("data"):
                mime = blob.get("mimeType") or blob.get("mime_type")
                suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")
                save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))

save_result(resp.to_dict(mode="json"))
```

```js
import OpenAI from "openai";

if (!process.env.AILOOT_API_KEY?.trim()) throw new Error("Set AILOOT_API_KEY before making a request");
const client = new OpenAI({ baseURL: "https://api.ailoot.ai/v1", apiKey: process.env.AILOOT_API_KEY, maxRetries: 0 });
const resp = await client.chat.completions.create({
  model: "nvidia/llama-3.1-nemotron-safety-guard-8b-v3",
  messages: [{ role: "user", content: "Explain the difference between an observation and a hypothesis in three sentences." }],
});
import { writeFileSync } from "node:fs";

function saveResult(data, stem = "result") {
  writeFileSync(stem + ".json", JSON.stringify(data, null, 2));
  console.log(JSON.stringify(data, null, 2));
  const saveBlob = (encoded, suffix, label) => {
    const raw = Buffer.from(encoded, "base64");
    const ext = raw.subarray(0,8).equals(Buffer.from([137,80,78,71,13,10,26,10])) ? "png" : raw[0]===255 && raw[1]===216 ? "jpg" : raw.subarray(0,4).toString()==="RIFF" && raw.subarray(8,12).toString()==="WEBP" ? "webp" : suffix;
    const file = `${stem}-${label}.${ext}`;
    writeFileSync(file, raw); console.log("Saved", file);
  };
  for (const [i, block] of (data.content || []).entries()) console.log("Content", i, JSON.stringify(block));
  for (const choice of data.choices || []) console.log("Choice", choice.index, JSON.stringify(choice.message || choice.delta || {}));
  for (const [i, item] of (data.output || []).entries()) {
    console.log("Output", i, JSON.stringify(item));
    if (item.type === "image_generation_call" && item.result) saveBlob(item.result, item.output_format || "bin", `output-${i}`);
  }
  for (const [ci, candidate] of (data.candidates || []).entries()) {
    for (const [pi, part] of (candidate.content?.parts || []).entries()) {
      console.log("Candidate/part", ci, pi, JSON.stringify(part));
      const blob = part.inlineData || part.inline_data;
      if (blob?.data) {
        const mime = blob.mimeType || blob.mime_type;
        const suffix = ({"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"})[mime] || "bin";
        saveBlob(blob.data, suffix, `candidate-${ci}-part-${pi}`);
      }
    }
  }
}

saveResult(resp);
```

```python
from openai import OpenAI
import os
from pathlib import Path

if not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")
client=OpenAI(base_url="https://api.ailoot.ai/v1",api_key=os.environ["AILOOT_API_KEY"],max_retries=0)
import base64, json
from pathlib import Path

def save_result(data, stem="result"):
    # Keep every field, including usage, IDs, reasoning and tool requests.
    Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
    print(json.dumps(data, ensure_ascii=False, indent=2))
    def save_blob(encoded, suffix, label):
        raw = base64.b64decode(encoded, altchars=b"-_", validate=True)
        ext = "png" if raw.startswith(b"\x89PNG\r\n\x1a\n") else "jpg" if raw.startswith(b"\xff\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix
        file = Path(stem + "-" + label + "." + ext)
        file.write_bytes(raw)
        print("Saved", file)
    for index, block in enumerate(data.get("content") or []):
        print("Content", index, json.dumps(block, ensure_ascii=False))
    for choice in data.get("choices") or []:
        print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))
    for index, item in enumerate(data.get("output") or []):
        print("Output", index, json.dumps(item, ensure_ascii=False))
        if item.get("type") == "image_generation_call" and item.get("result"):
            save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))
    for ci, candidate in enumerate(data.get("candidates") or []):
        for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):
            print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))
            blob = part.get("inlineData") or part.get("inline_data")
            if blob and blob.get("data"):
                mime = blob.get("mimeType") or blob.get("mime_type")
                suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")
                save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))

with client.chat.completions.stream(model='nvidia/llama-3.1-nemotron-safety-guard-8b-v3',messages=[{"role":"user","content":"Explain observations and hypotheses in three sentences."}]) as stream:
    with Path("stream-events.jsonl").open("w",encoding="utf-8") as log:
        for event in stream:
            payload=event.to_dict(mode="json")
            log.write(json.dumps(payload,ensure_ascii=False)+"\n"); log.flush()
            print(json.dumps(payload,ensure_ascii=False))
    final=stream.get_final_completion()
    save_result(final.to_dict(mode="json"),"stream-result")
```

### Terima dan simpan hasil

#### Ailoot · OpenAI Chat

json atau SSE

##### Simpan semua hasil yang selesai

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

Simpan setiap alternatif jawaban beserta konten, panggilan alat, dan penalarannya. Dalam SSE, gabungkan delta menurut indeks alternatif jawaban/alat, lalu pertahankan alasan penghentian akhir dan data penggunaan.

##### Jika respons asli perlu diperiksa

Simpan ID permintaan/respons dan header. URL umum untuk pengambilan tugas teks tidak disediakan.

Kolom respons

| Kolom | Jenis |
| --- | --- |
| Kolom`choices[].index` | Jenisinteger |
| Kolom`choices[].message.content` | Jenisstring\|null |
| Kolom`choices[].message.tool_calls` | Jenistool\_call\[\] |
| Kolom`choices[].message.reasoning_content` | Jenisstring |
| Kolom`choices[].message.reasoning_details` | Jenisobject\|array |
| Kolom`choices[].finish_reason` | Jenisstring\|null |
| Kolom`usage` | Jenisobject |

Kesalahan dan langkah berikutnya

| Respons | Kondisi | Langkah berikutnya |
| --- | --- | --- |
| Respons400 · `invalid_request_error` | KondisiBidang wajib, representasi materi, atau nilai parameter tidak valid | Langkah berikutnyaPerbaiki bidang yang ditunjukkan sebelum permintaan pembuatan baru. |

### Alias

#### Alias model

**Ailoot** [`nvidia/llama-3.1-nemotron-safety-guard-8b-v3`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&fmt=chat), [`nemotron-safety-guard`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard&fmt=chat), [`nemotron-safety-guard-latest`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard-latest&fmt=chat)

#### Alias jalur metode

Gunakan jalur metode yang ditampilkan di atas.

#### ID pemanggil terkait

[`nvidia/llama-3.1-nemotron-safety-guard-8b-v3`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&fmt=chat), [`nemotron-safety-guard`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard&fmt=chat), [`nemotron-safety-guard-latest`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard-latest&fmt=chat)

**Ailoot · OpenAI Chat** `POST /v1/chat/completions` `nemotron-safety-guard`

[`nemotron-safety-guard`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard&fmt=chat)

NVIDIA /`nemotron-safety-guard`

ID model

- `nvidia/llama-3.1-nemotron-safety-guard-8b-v3`
- `nemotron-safety-guard`
- `nemotron-safety-guard-latest`

### Mulai

Nemotron Safety Guard v3 memeriksa pesan pengguna dan jawaban model berdasarkan taksonomi risiko yang ditentukan. Mengembalikan JSON dengan penilaian keselamatan terpisah dan kategori pelanggaran.

Masukan yang diterima

Teks

Hasil

Teks

### Tersedia di Ailoot · masukan & metode

**Ailoot · OpenAI Chat** `POST /v1/chat/completions`

Masukan

`text`

Keluaran

`text`

Konteks

4.096 tokens

Keluaran maksimum

1.024 tokens

Format keluaran

json atau SSE

| Kolom | Jenis | Persyaratan | Nilai & batasan | Bawaan |
| --- | --- | --- | --- | --- |
| Kolom`model` | Jenis`string` | PersyaratanWajib | Nilai & batasan nemotron-safety-guard | Bawaan— |
| Kolom`messages` | Jenis`message[]` | PersyaratanWajib | Nilai & batasan Item minimum: 1 | Bawaan— |
| Kolom`messages[].role` | Jenis`string` | PersyaratanWajib | Nilai & batasan Contoh: system, user, assistant, tool, function Kondisi: Peran ini direpresentasikan secara eksplisit oleh tipe kompatibilitas dalam kode sumber; kumpulan peran model yang tepat tetap khusus untuk tiap model. Bergantung pada model | Bawaan— |
| Kolom`messages[].content` | Jenis`string\|content_part[]\|null` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`messages[].tool_calls` | Jenis`tool_call[]` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`messages[].tool_call_id` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Berlaku jika: role=tool; gunakan ID yang benar-benar dikembalikan | Bawaan— |
| Kolom`messages[].name` | Jenis`string` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`messages[].reasoning_content` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Berlaku jika: Kirim kembali penalaran yang benar-benar dikembalikan jika model ini menggunakannya | Bawaan— |
| Kolom`messages[].reasoning_details` | Jenis`object\|array` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`max_tokens` | Jenis`integer` | PersyaratanOpsional | Nilai & batasan Minimum: 1 Bergantung pada model | Bawaan— |
| Kolom`max_completion_tokens` | Jenis`integer` | PersyaratanOpsional | Nilai & batasan Minimum: 1 Bergantung pada model | Bawaan— |
| Kolom`stream` | Jenis`boolean` | PersyaratanOpsional | Nilai & batasan | Bawaanfalse |
| Kolom`stream_options.include_usage` | Jenis`boolean` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`temperature` | Jenis`number` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`top_p` | Jenis`number` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`stop` | Jenis`string\|string[]` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`tools` | Jenis`function_tool[]` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`tools[].type` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`tools[].function.name` | Jenis`string` | PersyaratanWajib | Nilai & batasan Berlaku jika: alat fungsi disediakan | Bawaan— |
| Kolom`tools[].function.description` | Jenis`string` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`tools[].function.parameters` | Jenis`JSON_schema` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`tool_choice` | Jenis`string\|object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`parallel_tool_calls` | Jenis`boolean` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`response_format` | Jenis`object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`reasoning_effort` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`reasoning` | Jenis`object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`service_tier` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`instructions` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Kolom kompatibilitas | Bawaan— |
| Kolom`functions` | Jenis`function[]` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`function_call` | Jenis`string\|object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`prompt_cache_options` | Jenis`object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |

Kombinasi bahan masukan

**Teks** · Teks UTF8

`messages[].content`

Parameter metode dan siklus hidup

| Parameter | Kontrak |
| --- | --- |
| Parameter`material_types` | Kontraktext |
| Parameter`required` | Kontrakmodel, messages |
| Parameter`optional_controls` | Kontraknull |
| Parameter`content_type` | Kontrakapplication/json |
| Parameter`text_file_material` | Kontrakkonten UTF-8 aktual dimasukkan sebagai teks; ID file atau alamat URL (URLs) memerlukan dukungan yang tepat dan terpisah |

Siklus hidup: `exact retained published method`

Struktur metode dan ID bersifat eksak; kemampuan pemasok terpisah dari penerimaan permintaan.

### Paket

#### Tarif metode Ailoot

Masukan

US$0,0002 */ 1K*

Keluaran

US$0,0002 */ 1K*

Pembacaan cache

US$0,0002 */ 1K*

Penulisan cache

US$0,0002 */ 1K*

Penulisan cache, 5 menit

US$0,0002 */ 1K tokens*

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 →](https://ailoot.ai/docs/id/billing/reserve/)

### Contoh permintaan

#### Ailoot · OpenAI Chat

Contoh teksPython SDKNode.js SDKAliran Python

```bash
set -eu
: "${AILOOT_API_KEY:?Set your Ailoot API key}"
: "${AILOOT_TEXT_FILE:?Set your actual UTF-8 source-material file}"
export AILOOT_TEXT_FILE
python3 - <<'PY_REQUEST'
import json,os
from pathlib import Path
p=json.loads('{"model": "nemotron-safety-guard", "messages": []}')
f=Path(os.environ["AILOOT_TEXT_FILE"])
if not f.is_file():raise SystemExit("Choose your actual UTF-8 source material")
text=f.read_text(encoding="utf-8")
if not text.strip():raise SystemExit("Source material must contain text")
p["messages"]=[{"role":"user","content":"Summarize this material and quote supporting source passages.\n\n"+text}]
Path("request.json").write_text(json.dumps(p,ensure_ascii=False))
PY_REQUEST
curl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/chat/completions \
  -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \
  --data-binary @request.json -D response-headers.txt -o result.json
python3 - <<'PY_RESULT'
import json
from pathlib import Path
print(json.dumps(json.loads(Path("result.json").read_text()),ensure_ascii=False,indent=2))
PY_RESULT
```

```python
from openai import OpenAI
import os

if not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")
client = OpenAI(base_url="https://api.ailoot.ai/v1", api_key=os.environ["AILOOT_API_KEY"], max_retries=0)
resp = client.chat.completions.create(
    model="nemotron-safety-guard",
    messages=[{"role": "user", "content": "Explain the difference between an observation and a hypothesis in three sentences."}],
)
import base64, json
from pathlib import Path

def save_result(data, stem="result"):
    # Keep every field, including usage, IDs, reasoning and tool requests.
    Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
    print(json.dumps(data, ensure_ascii=False, indent=2))
    def save_blob(encoded, suffix, label):
        raw = base64.b64decode(encoded, altchars=b"-_", validate=True)
        ext = "png" if raw.startswith(b"\x89PNG\r\n\x1a\n") else "jpg" if raw.startswith(b"\xff\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix
        file = Path(stem + "-" + label + "." + ext)
        file.write_bytes(raw)
        print("Saved", file)
    for index, block in enumerate(data.get("content") or []):
        print("Content", index, json.dumps(block, ensure_ascii=False))
    for choice in data.get("choices") or []:
        print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))
    for index, item in enumerate(data.get("output") or []):
        print("Output", index, json.dumps(item, ensure_ascii=False))
        if item.get("type") == "image_generation_call" and item.get("result"):
            save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))
    for ci, candidate in enumerate(data.get("candidates") or []):
        for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):
            print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))
            blob = part.get("inlineData") or part.get("inline_data")
            if blob and blob.get("data"):
                mime = blob.get("mimeType") or blob.get("mime_type")
                suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")
                save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))

save_result(resp.to_dict(mode="json"))
```

```js
import OpenAI from "openai";

if (!process.env.AILOOT_API_KEY?.trim()) throw new Error("Set AILOOT_API_KEY before making a request");
const client = new OpenAI({ baseURL: "https://api.ailoot.ai/v1", apiKey: process.env.AILOOT_API_KEY, maxRetries: 0 });
const resp = await client.chat.completions.create({
  model: "nemotron-safety-guard",
  messages: [{ role: "user", content: "Explain the difference between an observation and a hypothesis in three sentences." }],
});
import { writeFileSync } from "node:fs";

function saveResult(data, stem = "result") {
  writeFileSync(stem + ".json", JSON.stringify(data, null, 2));
  console.log(JSON.stringify(data, null, 2));
  const saveBlob = (encoded, suffix, label) => {
    const raw = Buffer.from(encoded, "base64");
    const ext = raw.subarray(0,8).equals(Buffer.from([137,80,78,71,13,10,26,10])) ? "png" : raw[0]===255 && raw[1]===216 ? "jpg" : raw.subarray(0,4).toString()==="RIFF" && raw.subarray(8,12).toString()==="WEBP" ? "webp" : suffix;
    const file = `${stem}-${label}.${ext}`;
    writeFileSync(file, raw); console.log("Saved", file);
  };
  for (const [i, block] of (data.content || []).entries()) console.log("Content", i, JSON.stringify(block));
  for (const choice of data.choices || []) console.log("Choice", choice.index, JSON.stringify(choice.message || choice.delta || {}));
  for (const [i, item] of (data.output || []).entries()) {
    console.log("Output", i, JSON.stringify(item));
    if (item.type === "image_generation_call" && item.result) saveBlob(item.result, item.output_format || "bin", `output-${i}`);
  }
  for (const [ci, candidate] of (data.candidates || []).entries()) {
    for (const [pi, part] of (candidate.content?.parts || []).entries()) {
      console.log("Candidate/part", ci, pi, JSON.stringify(part));
      const blob = part.inlineData || part.inline_data;
      if (blob?.data) {
        const mime = blob.mimeType || blob.mime_type;
        const suffix = ({"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"})[mime] || "bin";
        saveBlob(blob.data, suffix, `candidate-${ci}-part-${pi}`);
      }
    }
  }
}

saveResult(resp);
```

```python
from openai import OpenAI
import os
from pathlib import Path

if not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")
client=OpenAI(base_url="https://api.ailoot.ai/v1",api_key=os.environ["AILOOT_API_KEY"],max_retries=0)
import base64, json
from pathlib import Path

def save_result(data, stem="result"):
    # Keep every field, including usage, IDs, reasoning and tool requests.
    Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
    print(json.dumps(data, ensure_ascii=False, indent=2))
    def save_blob(encoded, suffix, label):
        raw = base64.b64decode(encoded, altchars=b"-_", validate=True)
        ext = "png" if raw.startswith(b"\x89PNG\r\n\x1a\n") else "jpg" if raw.startswith(b"\xff\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix
        file = Path(stem + "-" + label + "." + ext)
        file.write_bytes(raw)
        print("Saved", file)
    for index, block in enumerate(data.get("content") or []):
        print("Content", index, json.dumps(block, ensure_ascii=False))
    for choice in data.get("choices") or []:
        print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))
    for index, item in enumerate(data.get("output") or []):
        print("Output", index, json.dumps(item, ensure_ascii=False))
        if item.get("type") == "image_generation_call" and item.get("result"):
            save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))
    for ci, candidate in enumerate(data.get("candidates") or []):
        for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):
            print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))
            blob = part.get("inlineData") or part.get("inline_data")
            if blob and blob.get("data"):
                mime = blob.get("mimeType") or blob.get("mime_type")
                suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")
                save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))

with client.chat.completions.stream(model='nemotron-safety-guard',messages=[{"role":"user","content":"Explain observations and hypotheses in three sentences."}]) as stream:
    with Path("stream-events.jsonl").open("w",encoding="utf-8") as log:
        for event in stream:
            payload=event.to_dict(mode="json")
            log.write(json.dumps(payload,ensure_ascii=False)+"\n"); log.flush()
            print(json.dumps(payload,ensure_ascii=False))
    final=stream.get_final_completion()
    save_result(final.to_dict(mode="json"),"stream-result")
```

### Terima dan simpan hasil

#### Ailoot · OpenAI Chat

json atau SSE

##### Simpan semua hasil yang selesai

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

Simpan setiap alternatif jawaban beserta konten, panggilan alat, dan penalarannya. Dalam SSE, gabungkan delta menurut indeks alternatif jawaban/alat, lalu pertahankan alasan penghentian akhir dan data penggunaan.

##### Jika respons asli perlu diperiksa

Simpan ID permintaan/respons dan header. URL umum untuk pengambilan tugas teks tidak disediakan.

Kolom respons

| Kolom | Jenis |
| --- | --- |
| Kolom`choices[].index` | Jenisinteger |
| Kolom`choices[].message.content` | Jenisstring\|null |
| Kolom`choices[].message.tool_calls` | Jenistool\_call\[\] |
| Kolom`choices[].message.reasoning_content` | Jenisstring |
| Kolom`choices[].message.reasoning_details` | Jenisobject\|array |
| Kolom`choices[].finish_reason` | Jenisstring\|null |
| Kolom`usage` | Jenisobject |

Kesalahan dan langkah berikutnya

| Respons | Kondisi | Langkah berikutnya |
| --- | --- | --- |
| Respons400 · `invalid_request_error` | KondisiBidang wajib, representasi materi, atau nilai parameter tidak valid | Langkah berikutnyaPerbaiki bidang yang ditunjukkan sebelum permintaan pembuatan baru. |

### Alias

#### Alias model

**Ailoot** [`nvidia/llama-3.1-nemotron-safety-guard-8b-v3`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&fmt=chat), [`nemotron-safety-guard`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard&fmt=chat), [`nemotron-safety-guard-latest`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard-latest&fmt=chat)

#### Alias jalur metode

Gunakan jalur metode yang ditampilkan di atas.

#### ID pemanggil terkait

[`nvidia/llama-3.1-nemotron-safety-guard-8b-v3`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&fmt=chat), [`nemotron-safety-guard`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard&fmt=chat), [`nemotron-safety-guard-latest`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard-latest&fmt=chat)

**Ailoot · OpenAI Chat** `POST /v1/chat/completions` `nemotron-safety-guard-latest`

[`nemotron-safety-guard-latest`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard-latest&fmt=chat)

NVIDIA /`nemotron-safety-guard-latest`

ID model

- `nvidia/llama-3.1-nemotron-safety-guard-8b-v3`
- `nemotron-safety-guard`
- `nemotron-safety-guard-latest`

### Mulai

Nemotron Safety Guard v3 memeriksa pesan pengguna dan jawaban model berdasarkan taksonomi risiko yang ditentukan. Mengembalikan JSON dengan penilaian keselamatan terpisah dan kategori pelanggaran.

Masukan yang diterima

Teks

Hasil

Teks

### Tersedia di Ailoot · masukan & metode

**Ailoot · OpenAI Chat** `POST /v1/chat/completions`

Masukan

`text`

Keluaran

`text`

Konteks

4.096 tokens

Keluaran maksimum

1.024 tokens

Format keluaran

json atau SSE

| Kolom | Jenis | Persyaratan | Nilai & batasan | Bawaan |
| --- | --- | --- | --- | --- |
| Kolom`model` | Jenis`string` | PersyaratanWajib | Nilai & batasan nemotron-safety-guard-latest | Bawaan— |
| Kolom`messages` | Jenis`message[]` | PersyaratanWajib | Nilai & batasan Item minimum: 1 | Bawaan— |
| Kolom`messages[].role` | Jenis`string` | PersyaratanWajib | Nilai & batasan Contoh: system, user, assistant, tool, function Kondisi: Peran ini direpresentasikan secara eksplisit oleh tipe kompatibilitas dalam kode sumber; kumpulan peran model yang tepat tetap khusus untuk tiap model. Bergantung pada model | Bawaan— |
| Kolom`messages[].content` | Jenis`string\|content_part[]\|null` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`messages[].tool_calls` | Jenis`tool_call[]` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`messages[].tool_call_id` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Berlaku jika: role=tool; gunakan ID yang benar-benar dikembalikan | Bawaan— |
| Kolom`messages[].name` | Jenis`string` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`messages[].reasoning_content` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Berlaku jika: Kirim kembali penalaran yang benar-benar dikembalikan jika model ini menggunakannya | Bawaan— |
| Kolom`messages[].reasoning_details` | Jenis`object\|array` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`max_tokens` | Jenis`integer` | PersyaratanOpsional | Nilai & batasan Minimum: 1 Bergantung pada model | Bawaan— |
| Kolom`max_completion_tokens` | Jenis`integer` | PersyaratanOpsional | Nilai & batasan Minimum: 1 Bergantung pada model | Bawaan— |
| Kolom`stream` | Jenis`boolean` | PersyaratanOpsional | Nilai & batasan | Bawaanfalse |
| Kolom`stream_options.include_usage` | Jenis`boolean` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`temperature` | Jenis`number` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`top_p` | Jenis`number` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`stop` | Jenis`string\|string[]` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`tools` | Jenis`function_tool[]` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`tools[].type` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`tools[].function.name` | Jenis`string` | PersyaratanWajib | Nilai & batasan Berlaku jika: alat fungsi disediakan | Bawaan— |
| Kolom`tools[].function.description` | Jenis`string` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`tools[].function.parameters` | Jenis`JSON_schema` | PersyaratanOpsional | Nilai & batasan | Bawaan— |
| Kolom`tool_choice` | Jenis`string\|object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`parallel_tool_calls` | Jenis`boolean` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`response_format` | Jenis`object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`reasoning_effort` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`reasoning` | Jenis`object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`service_tier` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`instructions` | Jenis`string` | PersyaratanOpsional | Nilai & batasan Kolom kompatibilitas | Bawaan— |
| Kolom`functions` | Jenis`function[]` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`function_call` | Jenis`string\|object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |
| Kolom`prompt_cache_options` | Jenis`object` | PersyaratanOpsional | Nilai & batasan Bergantung pada model | Bawaan— |

Kombinasi bahan masukan

**Teks** · Teks UTF8

`messages[].content`

Parameter metode dan siklus hidup

| Parameter | Kontrak |
| --- | --- |
| Parameter`material_types` | Kontraktext |
| Parameter`required` | Kontrakmodel, messages |
| Parameter`optional_controls` | Kontraknull |
| Parameter`content_type` | Kontrakapplication/json |
| Parameter`text_file_material` | Kontrakkonten UTF-8 aktual dimasukkan sebagai teks; ID file atau alamat URL (URLs) memerlukan dukungan yang tepat dan terpisah |

Siklus hidup: `exact retained published method`

Struktur metode dan ID bersifat eksak; kemampuan pemasok terpisah dari penerimaan permintaan.

### Paket

#### Tarif metode Ailoot

Masukan

US$0,0002 */ 1K*

Keluaran

US$0,0002 */ 1K*

Pembacaan cache

US$0,0002 */ 1K*

Penulisan cache

US$0,0002 */ 1K*

Penulisan cache, 5 menit

US$0,0002 */ 1K tokens*

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 →](https://ailoot.ai/docs/id/billing/reserve/)

### Contoh permintaan

#### Ailoot · OpenAI Chat

Contoh teksPython SDKNode.js SDKAliran Python

```bash
set -eu
: "${AILOOT_API_KEY:?Set your Ailoot API key}"
: "${AILOOT_TEXT_FILE:?Set your actual UTF-8 source-material file}"
export AILOOT_TEXT_FILE
python3 - <<'PY_REQUEST'
import json,os
from pathlib import Path
p=json.loads('{"model": "nemotron-safety-guard-latest", "messages": []}')
f=Path(os.environ["AILOOT_TEXT_FILE"])
if not f.is_file():raise SystemExit("Choose your actual UTF-8 source material")
text=f.read_text(encoding="utf-8")
if not text.strip():raise SystemExit("Source material must contain text")
p["messages"]=[{"role":"user","content":"Summarize this material and quote supporting source passages.\n\n"+text}]
Path("request.json").write_text(json.dumps(p,ensure_ascii=False))
PY_REQUEST
curl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/chat/completions \
  -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \
  --data-binary @request.json -D response-headers.txt -o result.json
python3 - <<'PY_RESULT'
import json
from pathlib import Path
print(json.dumps(json.loads(Path("result.json").read_text()),ensure_ascii=False,indent=2))
PY_RESULT
```

```python
from openai import OpenAI
import os

if not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")
client = OpenAI(base_url="https://api.ailoot.ai/v1", api_key=os.environ["AILOOT_API_KEY"], max_retries=0)
resp = client.chat.completions.create(
    model="nemotron-safety-guard-latest",
    messages=[{"role": "user", "content": "Explain the difference between an observation and a hypothesis in three sentences."}],
)
import base64, json
from pathlib import Path

def save_result(data, stem="result"):
    # Keep every field, including usage, IDs, reasoning and tool requests.
    Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
    print(json.dumps(data, ensure_ascii=False, indent=2))
    def save_blob(encoded, suffix, label):
        raw = base64.b64decode(encoded, altchars=b"-_", validate=True)
        ext = "png" if raw.startswith(b"\x89PNG\r\n\x1a\n") else "jpg" if raw.startswith(b"\xff\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix
        file = Path(stem + "-" + label + "." + ext)
        file.write_bytes(raw)
        print("Saved", file)
    for index, block in enumerate(data.get("content") or []):
        print("Content", index, json.dumps(block, ensure_ascii=False))
    for choice in data.get("choices") or []:
        print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))
    for index, item in enumerate(data.get("output") or []):
        print("Output", index, json.dumps(item, ensure_ascii=False))
        if item.get("type") == "image_generation_call" and item.get("result"):
            save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))
    for ci, candidate in enumerate(data.get("candidates") or []):
        for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):
            print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))
            blob = part.get("inlineData") or part.get("inline_data")
            if blob and blob.get("data"):
                mime = blob.get("mimeType") or blob.get("mime_type")
                suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")
                save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))

save_result(resp.to_dict(mode="json"))
```

```js
import OpenAI from "openai";

if (!process.env.AILOOT_API_KEY?.trim()) throw new Error("Set AILOOT_API_KEY before making a request");
const client = new OpenAI({ baseURL: "https://api.ailoot.ai/v1", apiKey: process.env.AILOOT_API_KEY, maxRetries: 0 });
const resp = await client.chat.completions.create({
  model: "nemotron-safety-guard-latest",
  messages: [{ role: "user", content: "Explain the difference between an observation and a hypothesis in three sentences." }],
});
import { writeFileSync } from "node:fs";

function saveResult(data, stem = "result") {
  writeFileSync(stem + ".json", JSON.stringify(data, null, 2));
  console.log(JSON.stringify(data, null, 2));
  const saveBlob = (encoded, suffix, label) => {
    const raw = Buffer.from(encoded, "base64");
    const ext = raw.subarray(0,8).equals(Buffer.from([137,80,78,71,13,10,26,10])) ? "png" : raw[0]===255 && raw[1]===216 ? "jpg" : raw.subarray(0,4).toString()==="RIFF" && raw.subarray(8,12).toString()==="WEBP" ? "webp" : suffix;
    const file = `${stem}-${label}.${ext}`;
    writeFileSync(file, raw); console.log("Saved", file);
  };
  for (const [i, block] of (data.content || []).entries()) console.log("Content", i, JSON.stringify(block));
  for (const choice of data.choices || []) console.log("Choice", choice.index, JSON.stringify(choice.message || choice.delta || {}));
  for (const [i, item] of (data.output || []).entries()) {
    console.log("Output", i, JSON.stringify(item));
    if (item.type === "image_generation_call" && item.result) saveBlob(item.result, item.output_format || "bin", `output-${i}`);
  }
  for (const [ci, candidate] of (data.candidates || []).entries()) {
    for (const [pi, part] of (candidate.content?.parts || []).entries()) {
      console.log("Candidate/part", ci, pi, JSON.stringify(part));
      const blob = part.inlineData || part.inline_data;
      if (blob?.data) {
        const mime = blob.mimeType || blob.mime_type;
        const suffix = ({"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"})[mime] || "bin";
        saveBlob(blob.data, suffix, `candidate-${ci}-part-${pi}`);
      }
    }
  }
}

saveResult(resp);
```

```python
from openai import OpenAI
import os
from pathlib import Path

if not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")
client=OpenAI(base_url="https://api.ailoot.ai/v1",api_key=os.environ["AILOOT_API_KEY"],max_retries=0)
import base64, json
from pathlib import Path

def save_result(data, stem="result"):
    # Keep every field, including usage, IDs, reasoning and tool requests.
    Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
    print(json.dumps(data, ensure_ascii=False, indent=2))
    def save_blob(encoded, suffix, label):
        raw = base64.b64decode(encoded, altchars=b"-_", validate=True)
        ext = "png" if raw.startswith(b"\x89PNG\r\n\x1a\n") else "jpg" if raw.startswith(b"\xff\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix
        file = Path(stem + "-" + label + "." + ext)
        file.write_bytes(raw)
        print("Saved", file)
    for index, block in enumerate(data.get("content") or []):
        print("Content", index, json.dumps(block, ensure_ascii=False))
    for choice in data.get("choices") or []:
        print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))
    for index, item in enumerate(data.get("output") or []):
        print("Output", index, json.dumps(item, ensure_ascii=False))
        if item.get("type") == "image_generation_call" and item.get("result"):
            save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))
    for ci, candidate in enumerate(data.get("candidates") or []):
        for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):
            print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))
            blob = part.get("inlineData") or part.get("inline_data")
            if blob and blob.get("data"):
                mime = blob.get("mimeType") or blob.get("mime_type")
                suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")
                save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))

with client.chat.completions.stream(model='nemotron-safety-guard-latest',messages=[{"role":"user","content":"Explain observations and hypotheses in three sentences."}]) as stream:
    with Path("stream-events.jsonl").open("w",encoding="utf-8") as log:
        for event in stream:
            payload=event.to_dict(mode="json")
            log.write(json.dumps(payload,ensure_ascii=False)+"\n"); log.flush()
            print(json.dumps(payload,ensure_ascii=False))
    final=stream.get_final_completion()
    save_result(final.to_dict(mode="json"),"stream-result")
```

### Terima dan simpan hasil

#### Ailoot · OpenAI Chat

json atau SSE

##### Simpan semua hasil yang selesai

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

Simpan setiap alternatif jawaban beserta konten, panggilan alat, dan penalarannya. Dalam SSE, gabungkan delta menurut indeks alternatif jawaban/alat, lalu pertahankan alasan penghentian akhir dan data penggunaan.

##### Jika respons asli perlu diperiksa

Simpan ID permintaan/respons dan header. URL umum untuk pengambilan tugas teks tidak disediakan.

Kolom respons

| Kolom | Jenis |
| --- | --- |
| Kolom`choices[].index` | Jenisinteger |
| Kolom`choices[].message.content` | Jenisstring\|null |
| Kolom`choices[].message.tool_calls` | Jenistool\_call\[\] |
| Kolom`choices[].message.reasoning_content` | Jenisstring |
| Kolom`choices[].message.reasoning_details` | Jenisobject\|array |
| Kolom`choices[].finish_reason` | Jenisstring\|null |
| Kolom`usage` | Jenisobject |

Kesalahan dan langkah berikutnya

| Respons | Kondisi | Langkah berikutnya |
| --- | --- | --- |
| Respons400 · `invalid_request_error` | KondisiBidang wajib, representasi materi, atau nilai parameter tidak valid | Langkah berikutnyaPerbaiki bidang yang ditunjukkan sebelum permintaan pembuatan baru. |

### Alias

#### Alias model

**Ailoot** [`nvidia/llama-3.1-nemotron-safety-guard-8b-v3`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&fmt=chat), [`nemotron-safety-guard`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard&fmt=chat), [`nemotron-safety-guard-latest`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard-latest&fmt=chat)

#### Alias jalur metode

Gunakan jalur metode yang ditampilkan di atas.

#### ID pemanggil terkait

[`nvidia/llama-3.1-nemotron-safety-guard-8b-v3`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&fmt=chat), [`nemotron-safety-guard`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard&fmt=chat), [`nemotron-safety-guard-latest`](https://ailoot.ai/models/id/#m=nvidia%2Fllama-3.1-nemotron-safety-guard-8b-v3&route=Core&caller=nemotron-safety-guard-latest&fmt=chat)

FAQ · nvidia/llama-3.1-nemotron-safety-guard-8b-v3

Input dan hasil apa yang tersedia?

`POST /v1/chat/completions` · `nvidia/llama-3.1-nemotron-safety-guard-8b-v3`: `text` → `text` `POST /v1/chat/completions` · `nemotron-safety-guard`: `text` → `text` `POST /v1/chat/completions` · `nemotron-safety-guard-latest`: `text` → `text`

Kolom apa yang digunakan permintaan?

`POST /v1/chat/completions` · `nvidia/llama-3.1-nemotron-safety-guard-8b-v3`: `model`, `messages`, `stream` `POST /v1/chat/completions` · `nemotron-safety-guard`: `model`, `messages`, `stream` `POST /v1/chat/completions` · `nemotron-safety-guard-latest`: `model`, `messages`, `stream`

Batasan apa yang berlaku?

`POST /v1/chat/completions` · `nvidia/llama-3.1-nemotron-safety-guard-8b-v3`: `ctx`: `4096`; `max_out`: `1024`; `messages.min_items`: `1` `POST /v1/chat/completions` · `nemotron-safety-guard`: `ctx`: `4096`; `max_out`: `1024`; `messages.min_items`: `1` `POST /v1/chat/completions` · `nemotron-safety-guard-latest`: `ctx`: `4096`; `max_out`: `1024`; `messages.min_items`: `1`

Hubungkan Ailoot ke Claude Code, OpenCode, Hermes, atau aplikasi Anda.

## Perusahaan dan portofolio proyeknya

Perusahaan dan portofolio proyeknya: jutaan pelanggan; penjualan miliaran setiap bulan.

## Beli token dengan diskon.

Diskon langganan untuk pekerjaan rutin. Kode promo — hingga 50% untuk pembelian token.

Langganan

Pilih paket dengan diskon untuk permintaan rutin Anda.

Kode promo hingga 50%

Masukkan kode saat membeli. Harga dihitung ulang sesuai promosi aktif.

Biaya dalam riwayat

Model, token, dan jumlah tagihan disimpan untuk setiap permintaan.

[Pilih langganan](https://ailoot.ai/docs/id/billing/subscriptions/) · [Lihat harga model](https://ailoot.ai/models/id/)

## Hubungkan API ke produk yang Anda jual.

SaaS, aplikasi, dan bot — sesuai penawaran dan perjanjian lisensi Ailoot. Untuk penjualan kembali, mintalah lisensi reseller.

Aplikasi komersial

Integrasikan pembuatan teks, gambar, video, dan suara dalam fitur berbayar layanan Anda.

Lisensi reseller

Kirim deskripsi proyek dan skema akses. Mintalah ketentuan penggunaan dan penjualan kembali secara tertulis.

Dokumen di situs

Penawaran, perjanjian lisensi, dan kebijakan pemrosesan data tersedia sebelum koneksi.

[Minta lisensi reseller](mailto:support@ailoot.ai) · [Penawaran dan perjanjian lisensi](https://ailoot.ai/legal/offer?lang=id)

## Ailoot · Perusahaan dan portofolio proyeknya





Ailoot bertumpu pada pengalaman operasional yang teruji, basis pelanggan yang besar, layanan yang stabil, dan pemantauan KPI berkelanjutan.





Perusahaan dan portofolio proyeknya: jutaan pelanggan; penjualan miliaran setiap bulan.





Ailoot memanfaatkan pengalaman operasional ini untuk akses API ke fitur AI dalam produk pelanggan, agen, aplikasi, dan otomatisasi.