Chat Completions

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

Isicelo esincinci yi-id yemodeli nomyalezo omnye womsebenzisi.

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."}]
  }'

Impendulo yi-JSON object enye:

{
  "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
  }
}

Ii-header

Ii-header zesicelo

Ii-header zempendulo

Iintsimi zesicelo

I-messages kuphela efunekayo. Ikholamu ethi Isebenza kwi ibiza iimodeli apho intsimi itshintsha impendulo. Iimodeli ze-open-weight ezisingathiweyo ziii-id ezilishumi elinambini zoluhlu lweemodeli; intsapho yeShannon 3 yi-shannon-3, shannon-3-pro, shannon-3.1 kunye ne-shannon-3.1-pro. Iimodeli namaxabiso

Ezinye iintsimi ze-OpenAI, ezifana ne-n, user, stream_options, parallel_tool_calls, presence_penalty, frequency_penalty, logit_bias, logprobs, metadata, store kunye ne-prompt_cache_key, ziyamkelwa ukuze ikhowudi ye-client ekhoyo isebenze ngaphandle kotshintsho. Azitshintshi mpendulo: ihlala ikho inketho enye, kwaye i-stream iphela ngokusetyenziswa.

Intsimi enohlobo olungalunganga lwe-JSON, umzekelo "max_tokens": "100", ibuyisa i-422. Isicelo esingenayo i-messages senza okufanayo.

Ii-tool, i-structured output, ukuqiqa nokufuna kwiwebhu nganye inephepha layo: Umsebenzi Ukufowuna, Iziphumo ezicwangcisiweyo, I-effort yokuqiqa, Uphendlo lwewebhu olwakhelwe ngaphakathi.

Isicelo esinezinketho

Esi sicelo sisetha umyalezo we-system, iintsimi zesampling ne-effort yokuqiqa. Sisebenzisa imodeli ye-open-weight esingathiweyo, ezisebenzisa zonke.

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"
  }'

Impendulo inemo efanayo nale ingentla. I-usage yayo yongeza iinkcukacha ezimbini kwiimodeli ze-open-weight ezisingathiweyo: ii-prompt token ezifundwe kwi-cache nee-token ezisetyenziswe ekuqiqeni.

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

Ubude be-output

I-max_tokens yenza izinto ezimbini. Kuqala, linani lee-token elibekelwe bucala kwibhalansi yakho xa isicelo siqala. Xa impendulo igqityiwe, loo mbeko ithatyathelwa indawo ngee-token isicelo esizisebenzisileyo. Ukuba i-max_tokens inkulu kunokushiyekileyo kwibhalansi yakho, isicelo sibuyisa i-429 Quota exceeded nokuba impendulo ngokwayo ibiya kungena. Thumela i-max_tokens ephantsi ukuze ubeke bucala okuncinci.

I-shannon-coder-1 ibalwa ngokwahlukileyo kule endpoint: isicelo ngasinye siyyenye yobizo lweShannon Coder lweplani yakho, kwaye akukho ii-token zibekelwa bucala. Imida nebhalansi

Okwesibini, ikhawulela ubude bempendulo kwezi modeli:

Ngaphandle kwe-max_tokens okanye i-max_completion_tokens, ixabiso li-4,096. Kwi-shannon-coder-1 li-65,536.

Imiyalezo

Umyalezo ngamnye yi-object ene-role ne-content. I-content yintambo yoonobumba, okanye uluhlu lweendawo xa umyalezo uthwele okungaphezu kombhalo.

Ngokusebenzisa i-id yentsapho yeShannon 3, faka imiyalelo ekufuneka ihlale ikhona kumyalezo we-user.

Kumanqanaba eShannon isicelo esingenambhalo womsebenzisi ne-tools sibuyisa i-400 No user message provided.

Iindawo zomxholo

Ubungakanani, imida noluhlu olupheleleyo lweefom zinephepha lazo. Imifanekiso neefayile

I-object yempendulo

Izizathu zokuphela

Impendulo engekho kwi-stream ixela i-stop okanye i-tool_calls.

Ukusetyenziswa

Kwiimodeli ze-open-weight ezisingathiweyo, i-prompt_tokens yimiyalezo yakho nenkcazo zee-tool ezibalwe nge-tokenizer yemodeli ngokwayo, kunye nee-token zemifanekiso. Ii-endpoint zokubala ii-token zibuyisa inani elifanayo phambi kokuba uthumele. Ukubala ii-token

Kumanqanaba eShannon, i-prompt_tokens ibala yonke into imodeli eyifundileyo ukuze ibhale impendulo, ngoko inkulu kunombhalo wemiyalezo yakho kuphela.

I-Streaming

Xa i-stream isetwe ku-true impendulo ifika njengeziganeko ze-chat.completion.chunk kwaye iphela nge-data: [DONE]. I-chunk yokugqibela phambi kwayo ithwala i-finish_reason ne-usage; akukho ntsimi ye-stream_options efunekayo. Iimo ze-chunk, imigca ye-keep-alive neempazamo ngaphakathi kwe-stream zinephepha lazo. Umsinga

Iimpazamo

Impazamo yi-JSON object enelungu le-error. Ukuhlolwa kuqhutywa ngolu landelelwano: i-API key, umzimba wesicelo, i-id yemodeli, emva koko ibhalansi. Itheyibhile idwelisa oko le endpoint ikubuyisa rhoqo. Uluhlu olupheleleyo, kunye nezinto ekufuneka uzame kwakhona, lunephepha lalo. Ukuphatha imposiso

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