Chat Completions

POST https://api.shannon-ai.com/v1/chat/completions

Mafi ƙanƙantar request shi ne model id da saƙon mai amfani ɗaya.

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.shannon-ai.com/v1",
)

response = client.chat.completions.create(
    model="shannon-3",
    messages=[{"role": "user", "content": "Say hello in one sentence."}],
)

print(response.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "YOUR_API_KEY",
  baseURL: "https://api.shannon-ai.com/v1",
});

const response = await client.chat.completions.create({
  model: "shannon-3",
  messages: [{ role: "user", content: "Say hello in one sentence." }],
});

console.log(response.choices[0].message.content);

curl https://api.shannon-ai.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "shannon-3",
    "messages": [{"role": "user", "content": "Say hello in one sentence."}]
  }'

Amsar JSON object ɗaya ce:

{
  "id": "chatcmpl-5f0c1e7a9b3d4c62a8e1f07d2b46c9a3",
  "object": "chat.completion",
  "created": 1791625200,
  "model": "shannon-3",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Hello, it is good to meet you.",
        "reasoning_content": "The user wants a greeting in one sentence. Keep it short and friendly."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 1184,
    "completion_tokens": 46,
    "total_tokens": 1230
  }
}

Headers

Headers na request

Headers na amsa

Fields na request

messages kaɗai ake buƙata. Ginshiƙin Wanda ke aiwatarwa yana sunayen models da field ke canza amsa a kansu. Hosted open-weight models su ne ids goma sha biyu na jerin models; iyalin Shannon 3 su ne shannon-3, shannon-3-pro, shannon-3.1 da shannon-3.1-pro. Models & farashi

Ana karɓar sauran fields na OpenAI, kamar n, user, stream_options, parallel_tool_calls, presence_penalty, frequency_penalty, logit_bias, logprobs, metadata, store da prompt_cache_key, don code na client da ke akwai ya yi aiki ba tare da canji ba. Ba sa canza amsar: koyaushe akwai choice ɗaya, kuma stream koyaushe yana ƙarewa da usage.

Field mai nau'in JSON mara kyau, misali "max_tokens": "100", yana mayar da 422. Request ba tare da messages ba ma haka.

Tools, structured output, reasoning da binciken yanar gizo kowanne yana da shafinsa: Kiran Aiki, Abubuwan da aka Tsara, Reasoning effort, Binciken Yanar Gizon da aka Gina.

Request mai zaɓuɓɓuka

Wannan request yana saita saƙon system, sampling fields da reasoning effort. Yana amfani da hosted open-weight model, wanda ke aiwatar da dukansu.

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.shannon-ai.com/v1",
)

response = client.chat.completions.create(
    model="DeepSeek-V4-Flash-0731-W4A16-AUTOROUND-REAP",
    messages=[
        {"role": "system", "content": "You are a physics teacher. Answer in two sentences."},
        {"role": "user", "content": "Why is the sky blue?"},
    ],
    max_tokens=512,
    temperature=0.3,
    top_p=0.9,
    seed=7,
    stop=["\n\n"],
    reasoning_effort="low",
)

message = response.choices[0].message
print(message.reasoning_content)  # the reasoning
print(message.content)            # the answer
print(response.usage)

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "YOUR_API_KEY",
  baseURL: "https://api.shannon-ai.com/v1",
});

const response = await client.chat.completions.create({
  model: "DeepSeek-V4-Flash-0731-W4A16-AUTOROUND-REAP",
  messages: [
    { role: "system", content: "You are a physics teacher. Answer in two sentences." },
    { role: "user", content: "Why is the sky blue?" },
  ],
  max_tokens: 512,
  temperature: 0.3,
  top_p: 0.9,
  seed: 7,
  stop: ["\n\n"],
  reasoning_effort: "low",
});

const message = response.choices[0].message;
console.log(message.reasoning_content); // the reasoning
console.log(message.content);           // the answer
console.log(response.usage);

curl https://api.shannon-ai.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "DeepSeek-V4-Flash-0731-W4A16-AUTOROUND-REAP",
    "messages": [
      {"role": "system", "content": "You are a physics teacher. Answer in two sentences."},
      {"role": "user", "content": "Why is the sky blue?"}
    ],
    "max_tokens": 512,
    "temperature": 0.3,
    "top_p": 0.9,
    "seed": 7,
    "stop": ["\n\n"],
    "reasoning_effort": "low"
  }'

Amsar tana da siffa iri ɗaya da ta sama. usage ɗinta yana ƙara bayanai biyu a hosted open-weight models: prompt tokens da aka karanta daga cache da tokens da aka kashe a reasoning.

{
  "usage": {
    "prompt_tokens": 31,
    "completion_tokens": 62,
    "total_tokens": 93,
    "prompt_tokens_details": {
      "cached_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 21
    }
  }
}

Tsawon output

max_tokens yana yin abubuwa biyu. Na farko, shi ne adadin tokens da ake ware daga balance ɗinku lokacin da request ya fara. Idan amsar ta cika, ana maye gurbin wannan adadi da tokens da request ya yi amfani da su. Idan max_tokens ya fi abin da ya rage a balance ɗinku girma, request yana mayar da 429 Quota exceeded ko da amsar da kanta za ta isa. Ku aika max_tokens mafi ƙanƙanta don ware ƙasa.

Ana ƙirga shannon-coder-1 daban a wannan endpoint: kowace request kira ɗaya ne na Shannon Coder na plan ɗinku, kuma ba a ware tokens don ita ba. Iyakoki da balance

Na biyu, yana iyakance tsawon amsa a waɗannan models:

Idan babu max_tokens ko max_completion_tokens, ƙimar ita ce 4,096. A shannon-coder-1 ita ce 65,536.

Saƙonni

Kowane saƙo object ne mai role da content. content string ne, ko array na sassa idan saƙon yana ɗauke da fiye da rubutu.

Da id na iyalin Shannon 3, ku saka umarnin da dole ya yi aiki a cikin saƙon user.

A matakan Shannon request ba tare da rubutun mai amfani ba kuma ba tare da tools ba yana mayar da 400 No user message provided.

Sassan abun ciki

Girma, iyakoki da cikakken jerin siffofi suna da shafinsu. Hotuna da fayiloli

Object na amsa

Dalilan ƙarewa

Amsar da ba a yi streaming ba tana bayyana stop ko tool_calls.

Usage

A hosted open-weight models, prompt_tokens saƙonninku ne da ma'anar tools da aka ƙirga da tokenizer na model da kansa, tare da tokens na kowane hoto. Endpoints na ƙirgan tokens suna mayar da lamba iri ɗaya kafin ku aika. Ƙidayar tokens

A matakan Shannon, prompt_tokens yana ƙirga duk abin da model ya karanta don rubuta amsar, don haka ya fi rubutun saƙonninku kaɗai girma.

Streaming

Idan an saita stream zuwa true amsar tana zuwa a matsayin events na chat.completion.chunk kuma tana ƙarewa da data: [DONE]. Chunk na ƙarshe kafin shi yana ɗauke da finish_reason da usage; ba a buƙatar stream_options. Siffofin chunks, layukan keep-alive da kurakurai a cikin stream suna da shafinsu. Yawo

Kurakurai

Kuskure JSON object ne mai member na error. Ana yin dubawa bisa wannan tsari: maɓallin API, jikin request, model id, sannan balance. Teburin yana lissafa abin da wannan endpoint ya fi mayarwa. Cikakken jerin, tare da abin da za a sake gwadawa, yana da shafinsa. Kuskuren Gudanarwa

{
  "error": {
    "type": "invalid_request_error",
    "message": "unknown model: no-such-model"
  }
}