Chat Completions

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

Petitio minima est id modelli et unus nuntius utentis.

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."}]
  }'

Responsum est unum objectum JSON:

{
  "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 petitionis

Headers responsi

Campi petitionis

Solum messages requiritur. Columna Applicatur a modella nominat in quibus campus responsum mutat. Modella open-weight hospitata sunt duodecim ids indicis modellorum; familia Shannon 3 est shannon-3, shannon-3-pro, shannon-3.1 et shannon-3.1-pro. Modella et pretia

Alii campi OpenAI, ut n, user, stream_options, parallel_tool_calls, presence_penalty, frequency_penalty, logit_bias, logprobs, metadata, store et prompt_cache_key, accipiuntur ut codex clientis existens immutatus currat. Responsum non mutant: semper una electio est, et fluxus semper cum usu finitur.

Campus cum genere JSON falso, exempli gratia "max_tokens": "100", 422 reddit. Petitio sine messages idem facit.

Instrumenta, output structus, ratiocinatio et quaestus interretialis paginam propriam quaeque habent: Vocatio functionum, Exitus structi, Conatus ratiocinationis, Quaestus interretialis.

Petitio cum optionibus

Haec petitio nuntium system, campos sampling et conatum ratiocinationis statuit. Modellum open-weight hospitatum utitur, quod omnia applicat.

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"
  }'

Responsum eandem formam habet ac supra. Eius usage in modellis open-weight hospitatis duo singula addit: tokens prompti e cache lectos et tokens in ratiocinationem impensos.

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

Longitudo outputi

max_tokens duo facit. Primum, est numerus tokenorum qui a saldo tuo reservatur cum petitio incipit. Cum responsum completum est, haec quantitas tokens quibus petitio usa est substituitur. Si max_tokens maius est quam quod de saldo tuo superest, petitio 429 Quota exceeded reddit etiam si responsum ipsum capi potuisset. max_tokens inferius mitte ut minus reserves.

shannon-coder-1 in hoc endpoint aliter computatur: quaeque petitio est una ex vocationibus Shannon Coder plani tui, et nulli tokens pro ea reservantur. Limites et saldum

Secundum, longitudinem responsi in his modellis limitat:

Sine max_tokens vel max_completion_tokens valor est 4,096. In shannon-coder-1 est 65,536.

Nuntii

Quisque nuntius est objectum cum role et content. content est series litterarum, vel ordo partium cum nuntius plus quam textum fert.

Cum id familiae Shannon 3, instructiones quae valere debent in nuntium user pone.

In gradibus Shannon petitio sine textu utentis et sine tools 400 No user message provided reddit.

Partes contenti

Magnitudines, limites et index plenus formarum paginam propriam habent. Imagines et fasciculi

Objectum responsi

Causae finis

Responsum non per fluxum missum stop vel tool_calls nuntiat.

Usus

In modellis open-weight hospitatis, prompt_tokens sunt nuntii tui et definitiones instrumentorum cum tokenizer proprio modelli computati, plus tokens quarumlibet imaginum. Endpoints computationis tokenorum eundem numerum reddunt antequam mittis. Computatio tokenum

In gradibus Shannon, prompt_tokens omnia computat quae modellum legit ut responsum scriberet, itaque maior est quam textus nuntiorum tuorum solus.

Fluxus

Cum stream in true positum est, responsum ut eventus chat.completion.chunk advenit et cum data: [DONE] finitur. Ultimum chunk ante id finish_reason et usage fert; nulla stream_options necessaria sunt. Formae chunk, lineae keep-alive et errores intra fluxum paginam propriam habent. Fluens

Errores

Error est objectum JSON cum membro error. Inspectiones hoc ordine fiunt: clavis API, corpus petitionis, id modelli, deinde saldum. Tabula ea enumerat quae hoc endpoint saepissime reddit. Index plenus, cum notis quid iterandum sit, paginam propriam habet. Tractatio errorum

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