Model-yo & qiimo

model aan la faragelin

22

daaqadda context-ka

256K

laga bilaabo

$0.50 /1M

Model-yada open-weight ee la martigeliyo

262,144-token context id kasta; input iyo output si gooni ah ayaa loo xisaabiyaa; id kasta wuxuu aqbalaa wixii model-ka uu ku dayanayo aqbalo

Model	Input	Structured output	Gudaha / cached / dibadda halkii 1M	Kuwo kale
DeepSeek-V4-Pro-08133BIT-REAPDeepSeek · 1.6T MoE · 49B activeAan la faragelin	Qoraal gelinaya	JSON schema	$1.95 / $0.488 / $3.90	Kuwo kale
GLM-5.23BIT-REAPZ.ai · 744B MoE · 40B activeAan la faragelin	Qoraal gelinaya	JSON schema	$0.73 / $0.183 / $2.34	Kuwo kale
Kimi-K33BIT-REAPMoonshot AI · 2.8T MoE · 104B activeAan la faragelin	Qoraal + sawir gelinaya	JSON schema	$3.83 / $0.958 / $19.12	Kuwo kale
Nemotron3Ultra3BIT-REAPNVIDIA · 550B hybrid Mamba-Attention MoE · 55B activeAan la faragelin	Qoraal gelinaya	JSON schema	$0.75 / $0.188 / $3.30	Kuwo kale
MiniMax-M33BIT-REAPMiniMax · 428B MoE · 23B active · sparse attentionAan la faragelin	Qoraal + sawir gelinaya	JSON schema	$0.50 / $0.125 / $2.00	Kuwo kale
DeepSeek-V4-Flash-0731W4A16-AUTOROUND-REAPDeepSeek · 284B MoE · 13B activeAan la faragelin	Qoraal gelinaya	JSON schema	$0.50 / $0.125 / $2.00	Kuwo kale
Kimi-K2.6W4A16-AUTOROUND-REAPMoonshot AI · 1T MoE · 32B activeAan la faragelin	Qoraal + sawir gelinaya	JSON schema	$0.78 / $0.195 / $3.67	Kuwo kale
Laguna-S-2.1W4A16-AUTOROUND-REAPPoolside · 118B MoE · 8B activeAan la faragelin	Qoraal gelinaya	Structured output ma jiro	$0.50 / $0.125 / $2.00	Kuwo kale
inklingW4A16-AUTOROUND-REAPThinking Machines · 975B MoE · 41B activeAan la faragelin	Qoraal + sawir gelinaya	Shay JSON	$1.42 / $0.355 / $6.07	Kuwo kale
MiMo-V2.5-ProW8A16Xiaomi · 1.02T MoE · 42B active · 8-bitAan la faragelin	Qoraal gelinaya	JSON schema	$0.50 / $0.125 / $2.00	Kuwo kale
MiMo-V2.5W8A16Xiaomi · 8-bit multimodal · W8A16Aan la faragelin	Qoraal + sawir gelinaya	JSON schema	$0.50 / $0.125 / $2.00	Kuwo kale
Hy3W8A16Tencent · 295B MoE · 21B active · 8-bitAan la faragelin	Qoraal gelinaya	JSON schema	$0.50 / $0.125 / $2.00	Kuwo kale

Model-yada Shannon

hal qiime go'an input-ka iyo output-ka

Model	Context	Input	Structured output	Qiimaha / 1M token
Shannon 1.6 Liteshannon-1.6-liteAan la faragelin	192K	Qoraal + sawir gelinaya	JSON schema	$3.90
Shannon 1.6 Proshannon-1.6-proAan la faragelin	192K	Qoraal + sawir gelinaya	JSON schema	$7.80
Shannon 2 Liteshannon-2-liteAan la faragelin	192K	Qoraal + sawir gelinaya	JSON schema	$3.90
Shannon 2 Proshannon-2-proAan la faragelin	192K	Qoraal + sawir gelinaya	JSON schema	$5.85
Shannon 3shannon-3Aan la faragelin	192K	Qoraal + sawir gelinaya	JSON schema	$3.35
Shannon 3 Proshannon-3-proAan la faragelin	192K	Qoraal + sawir gelinaya	JSON schema	$3.35
Shannon 3.1shannon-3.1Aan la faragelin	192K	Qoraal + sawir gelinaya	JSON schema	$3.35
Shannon 3.1 Proshannon-3.1-proAan la faragelin	192K	Qoraal + sawir gelinaya	JSON schema	$3.35
Shannon 3.5shannon-3.5Aan la faragelin	192K	Qoraal + sawir gelinaya	JSON schema	$3.35
Shannon Coder 1shannon-coder-1Aan la faragelin	128K	Qoraal gelinaya	JSON schema	$8.00

U gudub Bilawga Degdegga ah Isku day garoonka ciyaaraha

Sida safka loo akhriyo

Model kasta wuu stream-gareeyaa wuuna wici karaa tools. Astaamaha kale ee API-ga waxaa lagu sharaxay bogga Awoodaha. Awoodaha

Context window iyo xadka output-ka

• Context window-ku wuxuu daboolaa wax kasta oo model-ku akhriyo hal jawaab: fariimahaaga, qeexidaha tool iyo sawirrada.
• Qoyska Shannon 3 wada-hadal dheer isagaa isku habeeya: wuxuu hayaa wareegyada ugu cusub ee daaqadda ku soo gala wuuna jawaabaa.
• Stream dhexdiisa, prompt aad u dheer waxaa loo sheegaa sida frame khalad ah oo leh nooca invalid_request_error iyo code: "context_length_exceeded".
• max_tokens wuxuu qaataa qiimayaal 1 ilaa 65,536 model kasta; max_output_tokens ee GET /v1/models waa dhammaadka sare ee kala duwanaanshahaas. Default-ku waa 4,096.
• Model kasta, max_tokens waa cadadka codsi haraaggaaga ka keydiyo.
• Sida xadka dhererka jawaabta, max_tokens waxaa dabaqa model-yada open-weight ee la martigeliyo, shannon-1.6-lite, shannon-1.6-pro iyo shannon-coder-1.
• Model-yada open-weight ee la martigeliyo wuxuu xadidaa qoraalka jawaabta. Reasoning-ka jawaabta ka horreeya looma tiriyo, qiime ka hooseeya 256 -na wuxuu u shaqeeyaa sida 256.
• Stream shannon-1.6-lite, shannon-1.6-pro, shannon-coder-1 ama qoyska Shannon 3 wuxuu ku dhammaadaa finish_reason: "length" marka jawaabta xadkeeda lagu gooyo.

Xadad iyo haraag

Sida qiime loo sameeyo

• Qiimayaashu waa USD halkii 1,000,000 token. Haraaggaaga waxaa lagu tiriyaa token-yo qiimahoodu yahay $5.00 halkii 1,000,000, sidaas darteed qiimaha model-ku wuxuu sheegaa inta haraag ah ee token-yadiisu qaadato.
• Model-yada Shannon waxay leeyihiin hal qiime. Wicitaan waxaa lagu dallacaa usage.total_tokens qiimahaas.
• Model-yada open-weight ee la martigeliyo waxay leeyihiin saddex qiime. Input laga akhriyay prompt cache waxaa lagu dallacaa qiimaha cached, kaas oo ah 25% qiimaha input-ka oo lagu wareejiyay $0.001. Inta kale ee input-ka waxaa lagu dallacaa qiimaha input-ka, output-ka qiimaha output-ka.
• Model-yada open-weight ee la martigeliyo token-yada waxay ku tiriyaan tokenizer-ka model kasta. GET /v1/models waxay tan ku calaamadeeyaan tokenizer: "original".
• shannon-coder-1 waxaa lagu tiriyaa wicitaanno, ma aha token-yo, /v1/chat/completions iyo /v1/messages. Qiimihiisa la liis gareeyay wuxuu khuseeyaa /v1/responses.

Waxa codsi keydiyo, tusaalayaal dallac iyo allowance-ka wicitaannada Shannon Coder waxay ku yaallaan Xadad iyo haraag. Xadad iyo haraag

Id-yada model-ka

Dir mid ka mid ah id-yadan sida model.

shannon-1.6-lite
shannon-1.6-pro
shannon-2-lite
shannon-2-pro
shannon-3
shannon-3-pro
shannon-3.1
shannon-3.1-pro
shannon-3.5
shannon-coder-1
DeepSeek-V4-Pro-0813-3BIT-REAP
GLM-5.2-3BIT-REAP
Kimi-K3-3BIT-REAP
Nemotron3Ultra-3BIT-REAP
MiniMax-M3-3BIT-REAP
DeepSeek-V4-Flash-0731-W4A16-AUTOROUND-REAP
Kimi-K2.6-W4A16-AUTOROUND-REAP
Laguna-S-2.1-W4A16-AUTOROUND-REAP
inkling-W4A16-AUTOROUND-REAP
MiMo-V2.5-Pro-W8A16
MiMo-V2.5-W8A16
Hy3-W8A16

• Xarfaha waaweyn iyo yaryar muhiim ma aha, meelaha bannaan ee id-ga agtiisana waa la saaraa: SHANNON-3 iyo shannon-3 waa isla model.
• Jawaabtu had iyo jeer waxay model ku sidaa id-ga la daabacay, higgaad kasta oo aad dirtay ha ahaatee.
• Magac muuqaal sida Shannon 3 ma aha id.

Afar higgaad oo ka gaaban ayaa sidoo kale la aqbalaa.

Id aan la daabacin waxaa lagu jawaabaa xaalad 400, nooca invalid_request_error iyo fariinta unknown model: <id> saddexda endpoint. Model kale kaalinteeda ma jawaabo.

{
  "error": {
    "type": "invalid_request_error",
    "message": "unknown model: gpt-4o"
  }
}

{
  "type": "error",
  "error": {
    "type": "invalid_request_error",
    "message": "unknown model: gpt-4o"
  }
}

• Had iyo jeer dir model. /v1/responses wuxuu codsi aan lahayn ku jawaabaa 400. /v1/chat/completions iyo /v1/messages codsi sidaas ah wuxuu ka jawaabaa shannon-1.6-lite.
• POST /v1/tokenize iyo POST /v1/messages/count_tokens waxay qaataan id-yada open-weight ee la martigeliyo.

Liis model-yada

GET /v1/models wuxuu soo celiyaa model-yada bogga sida JSON, isla context windows, qiimayaal iyo awoodo. Ku isticmaal inaad buuxiso model picker ama inaad qiime ku hubiso code-ka.

GET https://api.shannon-ai.com/v1/models

from openai import OpenAI

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

for model in client.models.list():
    print(model.id)

import OpenAI from 'openai';

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

for await (const model of client.models.list()) {
  console.log(model.id);
}

curl "https://api.shannon-ai.com/v1/models"

Jawaabta, oo loo soo koobay hal gelin nooc kasta:

{
  "object": "list",
  "data": [
    {
      "id": "shannon-3",
      "object": "model",
      "created": 1700000000,
      "owned_by": "shannon-ai",
      "display_name": "Shannon 3",
      "context_window": 196608,
      "max_output_tokens": 65536,
      "pricing": {
        "input_per_million_usd": 3.35,
        "output_per_million_usd": 3.35,
        "currency": "USD"
      },
      "capabilities": {
        "tools": true,
        "vision": true,
        "reasoning": true,
        "reasoning_effort": false,
        "streaming": true,
        "json_schema": true
      },
      "endpoints": [
        "/v1/chat/completions",
        "/v1/messages",
        "/v1/responses"
      ]
    },
    {
      "id": "MiniMax-M3-3BIT-REAP",
      "object": "model",
      "created": 1700000000,
      "owned_by": "shannon-ai",
      "display_name": "MiniMax-M3 · 3BIT-REAP",
      "context_window": 262144,
      "max_output_tokens": 65536,
      "pricing": {
        "input_per_million_usd": 0.5,
        "cached_input_per_million_usd": 0.125,
        "output_per_million_usd": 2,
        "currency": "USD"
      },
      "tokenizer": "original",
      "capabilities": {
        "tools": true,
        "vision": true,
        "reasoning": true,
        "reasoning_effort": true,
        "streaming": true,
        "response_format": true,
        "json_schema": true,
        "prompt_caching": true
      },
      "endpoints": [
        "/v1/chat/completions",
        "/v1/messages",
        "/v1/responses",
        "/v1/messages/count_tokens",
        "/v1/tokenize"
      ]
    }
  ]
}

• GET /v1/models fure ma u baahna. Jawaabtu waxay isku mid tahay wacayaal kasta mana ku xirna qorshahaaga.
• Liisku wuxuu haystaa id-yada la wici karo. Model hadda la bixin karin liiska waa laga reebaa.
• Endpoint hal model ah ma jiro: GET /v1/models/{id} wuxuu ku jawaabaa 404 oo leh nooca not_found_error.