How do I set a maximum amount of money the API could cost me?

The API does not support hard spending caps directly. Monitor usage via the response headers or track tokens in your application logic. Prepaid credit never expires, so you control the limit by your top-up amount.

What happens if I hit the rate limit?

You receive a 429 status code. The request consumes no tokens. Wait a few seconds and retry. You are limited to 300 requests per minute per key. Adjust your concurrency to stay under 8 simultaneous requests.

Is the uncensored model the same as Mistral AI's models?

No. We host an open-weight model tuned for no refusals. It is not GPT, Claude, or the official Mistral models. It is compatible with the same API structure but has its own model ID. Check the pricing page for details.

Pengenalan: Integrasikan API Tanpa Sensor

URL Asas dan Pengesahan

API tinggal di https://api.mistralapipricing.com/v1. Lulus kunci anda dalam Authorization header. Logik harga API Mistral dikenakan setiap token, jadi pastikan header bersih. Sahkan sambungan dengan GET /v1/models.

Header: Authorization: Bearer YOUR_KEY
URL Asas: https://api.mistralapipricing.com/v1

curl https://api.mistralapipricing.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "List the cost per token for each model."}]
  }'

Menghantar Permintaan Pertama Anda

Panggil chat/completions endpoint dengan ID model anda. Model tanpa sensor menanggapi prompt standard tanpa penolakan. Token input berharga $0.15 per juta. Token output berharga $0.50 per juta. Ralat tidak menggunakan kredit.

from openai import OpenAI

client = OpenAI(base_url="https://api.mistralapipricing.com/v1", api_key="YOUR_KEY")

resp = client.chat.completions.create(
    model="uncensored",
    messages=[{"role": "user", "content": "Summarise this thread without softening it."}],
)
print(resp.choices[0].message.content)

Integrasi Node.js

Gunakan SDK rasmi atau axios untuk menghantar permintaan. Tetapkan URL asas dengan betul. ID model sentiasa uncensoredPengepala mesti termasuk token bearer anda. Respons mengembalikan objek JSON. Tangani ralat dengan baik.

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://api.mistralapipricing.com/v1", apiKey: process.env.API_KEY });

const resp = await client.chat.completions.create({
  model: "uncensored",
  messages: [{ role: "user", content: "Draft a villain monologue for my game." }],
});
console.log(resp.choices[0].message.content);

Respons Streaming

Aktifkan aliran dengan menetapkan stream: true. Terima token melalui Acara Dihantar Pelayan. Bahagian akhir mengandungi data penggunaan. Ini mengurangkan latensi yang dirasakan untuk output panjang. Buffer token di sisi klien. Paparkan mereka apabila mereka tiba.

stream = client.chat.completions.create(
    model="uncensored",
    messages=[{"role": "user", "content": "Tell the story in second person."}],
    stream=True,
)
for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

Panggilan Fungsi dan Mod JSON

Takrifkan alat dalam badan permintaan. Model mengembalikan JSON terstruktur. Gunakan response_format: {"type": "json_object"} untuk output ketat. Uraikan keputusan dalam logik aplikasi anda. Ini berfungsi untuk tugas automasi. Sahkan skema di hujung anda.

Had, Ralat, dan Konteks

Jendela konteks ialah 100,000 token secara keseluruhan. Permintaan dihadkan kepada 300 setiap minit. Anda mendapat 8 permintaan serentak. Ralat 401 bermaksud kunci tidak sah. Ralat 402 menunjukkan kredit tidak mencukupi. Ralat 429 menandakan had kadar. Output maksimum ialah 16,000 token.

Panduan Mula Pantas