Chat Completions

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

L-iżgħar talba hija id ta' mudell u messaġġ wieħed ta' utent.

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."}]
  }'

It-tweġiba hija oġġett JSON wieħed:

{
  "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 tat-talba

Headers tat-tweġiba

Kampi tat-talba

Hemm bżonn biss ta' messages. Il-kolonna Applikat minn issemmi l-mudelli li fuqhom qasam ibiddel it-tweġiba. Il-mudelli open-weight ospitati huma tnax-il id tal-lista tal-mudelli; il-familja Shannon 3 hija shannon-3, shannon-3-pro, shannon-3.1 u shannon-3.1-pro. Mudelli u prezzijiet

Kampi oħra ta' OpenAI, bħal n, user, stream_options, parallel_tool_calls, presence_penalty, frequency_penalty, logit_bias, logprobs, metadata, store u prompt_cache_key, jiġu aċċettati sabiex il-kodiċi eżistenti tal-klijent jaħdem mingħajr bidla. Ma jbiddlux it-tweġiba: dejjem hemm għażla waħda, u stream dejjem jispiċċa bl-użu.

Qasam bit-tip JSON ħażin, pereżempju "max_tokens": "100", jagħti lura 422. Hekk jagħmel ukoll talba mingħajr messages.

Tools, output strutturat, raġunament u fittxija fuq il-web kull wieħed għandu paġna tiegħu: Sejħiet ta' Funzjonijiet, Outputs Strutturati, Sforz ta' raġunament, Tfittxija Web inkorporata.

Talba b'għażliet

Din it-talba tistabbilixxi messaġġ system, il-kampi tas-sampling u l-isforz tar-raġunament. Tuża mudell open-weight ospitat, li japplika lil kollha.

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"
  }'

It-tweġiba għandha l-istess forma bħal hawn fuq. L-usage tagħha jżid żewġ dettalji fuq il-mudelli open-weight ospitati: il-prompt tokens moqrija mill-cache u t-tokens użati għar-raġunament.

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

Tul tal-output

max_tokens jagħmel żewġ affarijiet. L-ewwel, huwa n-numru ta' tokens li jitwarrbu mill-bilanċ tiegħek meta tibda t-talba. Meta t-tweġiba tkun kompluta, dak l-ammont jinbidel bit-tokens li t-talba użat. Jekk max_tokens huwa akbar minn dak li fadal mill-bilanċ tiegħek, it-talba tagħti lura 429 Quota exceeded anki jekk it-tweġiba nnifisha kienet taqbel. Ibgħat max_tokens aktar baxx biex tiwarrab inqas.

shannon-coder-1 jingħadd b'mod differenti fuq dan l-endpoint: kull talba hija waħda mis-sejħiet ta' Shannon Coder tal-pjan tiegħek, u ebda tokens ma jitwarrbu għaliha. Limiti u bilanċ

It-tieni, jillimita t-tul tat-tweġiba fuq dawn il-mudelli:

Mingħajr max_tokens jew max_completion_tokens, il-valur huwa 4,096. Fuq shannon-coder-1 huwa 65,536.

Messaġġi

Kull messaġġ huwa oġġett b'role u content. content huwa string, jew array ta' partijiet meta l-messaġġ iġorr aktar minn test.

B'id tal-familja Shannon 3, poġġi fil-messaġġ user l-istruzzjonijiet li jridu jinżammu.

Fuq il-livelli Shannon talba mingħajr test ta' utent u mingħajr tools tagħti lura 400 No user message provided.

Partijiet tal-kontenut

Id-daqsijiet, il-limiti u l-lista sħiħa tal-forom għandhom paġna tagħhom. Immaġini u fajls

L-oġġett tat-tweġiba

Raġunijiet tat-tmiem

Tweġiba mingħajr streaming tirrapporta stop jew tool_calls.

Użu

Fuq il-mudelli open-weight ospitati, prompt_tokens huma l-messaġġi u d-definizzjonijiet tat-tools tiegħek magħduda bit-tokenizer tal-mudell stess, flimkien mat-tokens ta' kwalunkwe immaġni. L-endpoints tal-għadd ta' tokens jagħtu lura l-istess numru qabel ma tibgħat. Għadd ta' tokens

Fuq il-livelli Shannon, prompt_tokens jgħodd dak kollu li l-mudell qara biex jikteb it-tweġiba, għalhekk huwa akbar mit-test tal-messaġġi tiegħek biss.

Streaming

B'stream issettjat għal true it-tweġiba tasal bħala avvenimenti chat.completion.chunk u tispiċċa b'data: [DONE]. L-aħħar chunk qablu jġorr finish_reason u usage; m'hemm bżonn ta' ebda stream_options. Il-forom tal-chunks, il-linji keep-alive u l-iżbalji ġo stream għandhom paġna tagħhom. Streaming

Żbalji

Żball huwa oġġett JSON b'membru error. Il-verifiki jmorru f'dan l-ordni: ċavetta tal-API, body tat-talba, id tal-mudell, imbagħad bilanċ. It-tabella telenka dak li dan l-endpoint jagħti lura l-aktar spiss. Il-lista kompluta, b'dak li għandu jerġa' jiġi pprovat, għandha paġna tagħha. Ġestjoni tal-iżbalji

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