Peke ki ngā ihirangi
Ngā model me ngā utu

Ngā model me ngā utu

MŌ IA 1M TOKENS

Ia id model o te API, ngā mea e whakaae ana ia me tōna utu: 10 model Shannon me 12 model open-weight hosted, e watea ana ia ki ngā endpoint e toru. Ka whakahoki te GET /v1/models i te rārangi ōrite hei JSON.

model kāore he aukati
22
matapihi horopaki
256K
mai i
$0.50 /1M

Ngā model open-weight hosted

Horopaki 262,144-token i ia id; ka whakawāteatia motuhake te input me te output; ka whakaae ia id ki tā te model e tāwhai ana e whakaae ai

ModelInputOutput whakatakotoI roto / cached / i waho per 1MĒtahi atu
DeepSeek-V4-Pro-08133BIT-REAPDeepSeek · 1.6T MoE · 49B activeKāore he aukatiKuputuhi ki rotoJSON schema$1.95 / $0.488 / $3.90
GLM-5.23BIT-REAPZ.ai · 744B MoE · 40B activeKāore he aukatiKuputuhi ki rotoJSON schema$0.73 / $0.183 / $2.34
Kimi-K33BIT-REAPMoonshot AI · 2.8T MoE · 104B activeKāore he aukatiKuputuhi + whakaahua ki rotoJSON schema$3.83 / $0.958 / $19.12
Nemotron3Ultra3BIT-REAPNVIDIA · 550B hybrid Mamba-Attention MoE · 55B activeKāore he aukatiKuputuhi ki rotoJSON schema$0.75 / $0.188 / $3.30
MiniMax-M33BIT-REAPMiniMax · 428B MoE · 23B active · sparse attentionKāore he aukatiKuputuhi + whakaahua ki rotoJSON schema$0.50 / $0.125 / $2.00
DeepSeek-V4-Flash-0731W4A16-AUTOROUND-REAPDeepSeek · 284B MoE · 13B activeKāore he aukatiKuputuhi ki rotoJSON schema$0.50 / $0.125 / $2.00
Kimi-K2.6W4A16-AUTOROUND-REAPMoonshot AI · 1T MoE · 32B activeKāore he aukatiKuputuhi + whakaahua ki rotoJSON schema$0.78 / $0.195 / $3.67
Laguna-S-2.1W4A16-AUTOROUND-REAPPoolside · 118B MoE · 8B activeKāore he aukatiKuputuhi ki rotoKāore he output whakatakoto$0.50 / $0.125 / $2.00
inklingW4A16-AUTOROUND-REAPThinking Machines · 975B MoE · 41B activeKāore he aukatiKuputuhi + whakaahua ki rotoJSON object$1.42 / $0.355 / $6.07
MiMo-V2.5-ProW8A16Xiaomi · 1.02T MoE · 42B active · 8-bitKāore he aukatiKuputuhi ki rotoJSON schema$0.50 / $0.125 / $2.00
MiMo-V2.5W8A16Xiaomi · 8-bit multimodal · W8A16Kāore he aukatiKuputuhi + whakaahua ki rotoJSON schema$0.50 / $0.125 / $2.00
Hy3W8A16Tencent · 295B MoE · 21B active · 8-bitKāore he aukatiKuputuhi ki rotoJSON schema$0.50 / $0.125 / $2.00

Ngā model Shannon

kotahi te utu paraihe mō te input me te output

ModelHoropakiInputOutput whakatakotoUtu / 1M tokens
Shannon 1.6 Liteshannon-1.6-liteKāore he aukati192KKuputuhi + whakaahua ki rotoJSON schema$3.90
Shannon 1.6 Proshannon-1.6-proKāore he aukati192KKuputuhi + whakaahua ki rotoJSON schema$7.80
Shannon 2 Liteshannon-2-liteKāore he aukati192KKuputuhi + whakaahua ki rotoJSON schema$3.90
Shannon 2 Proshannon-2-proKāore he aukati192KKuputuhi + whakaahua ki rotoJSON schema$5.85
Shannon 3shannon-3Kāore he aukati192KKuputuhi + whakaahua ki rotoJSON schema$3.35
Shannon 3 Proshannon-3-proKāore he aukati192KKuputuhi + whakaahua ki rotoJSON schema$3.35
Shannon 3.1shannon-3.1Kāore he aukati192KKuputuhi + whakaahua ki rotoJSON schema$3.35
Shannon 3.1 Proshannon-3.1-proKāore he aukati192KKuputuhi + whakaahua ki rotoJSON schema$3.35
Shannon 3.5shannon-3.5Kāore he aukati192KKuputuhi + whakaahua ki rotoJSON schema$3.35
Shannon Coder 1shannon-coder-1Kāore he aukati128KKuputuhi ki rotoJSON schema$8.00

Peke ki te Quickstart Whakamātau i te Playground

Te pānui i tētahi rārangi

Tīwae Te mea e kī ana
Model Te ingoa whakaatu. Ko te rārangi i raro iho ko te id ka tukuna e koe hei model. I te ripanga hosted ko te id te ingoa me te tohu i tōna taha, kua honoa mā te tohuhono: ka hanga te MiniMax-M3 me te 3BIT-REAP i te MiniMax-M3-3BIT-REAP. Ka whakaatu te rārangi i raro i te ingoa ko wai i hanga i te model me tōna rahi.
Horopaki Te matapihi horopaki: te maha o ngā token kōrero ka pānuitia e te model. Ko K te tohu mō te 1,024 token. Ko te tau tika ko context_window i te GET /v1/models.
Input Kuputuhi anake, kuputuhi me ngā whakaahua rānei. Ka pūrongo tētahi model e pānui ana i ngā whakaahua i te capabilities.vision: true. Ka pānui te shannon-2-lite me te shannon-2-pro i ngā whakaahua i ngā tono e tuku ana hoki i te tools i te response_format rānei.
Output whakaritea Te mea ka taea e te response_format te tono i te model: he whakautu e whai ana i tō JSON schema, he whakautu he ahanoa JSON, kāore rānei tētahi o aua mea.
Utu USD mō ia 1,000,000 token. Kotahi te utu a tētahi model Shannon mō te input me te output. E toru ā tētahi model open-weight hosted: input, input cached me output.

Ka stream ia model, ā, ka taea te karanga i ngā tool. Kua whakaahuatia ētahi atu āhuatanga o te API i te whārangi Capabilities. Āheinga

Te matapihi horopaki me ngā tepe output

  • Ka kapi te matapihi horopaki i ngā mea katoa ka pānui te model mō tētahi whakautu: ō karere, ngā whakamārama tool me ngā whakaahua.
  • Ka whakaūngia e te whānau Shannon 3 tētahi kōrero roa ake mā tōna ake: ka pupuri i ngā hurihanga hou rawa ka uru ki te matapihi, ā, ka whakautu.
  • I roto i tētahi stream, ka pūrongotia tētahi prompt roa rawa hei frame hapa me te momo invalid_request_error me te code: "context_length_exceeded".
  • Ka whakaae te max_tokens ki ngā uara mai i te 1 ki te 65,536 i ia model; ko te max_output_tokens i te GET /v1/models te tihi o taua whānuitanga. Ko te taunoa 4,096.
  • I ia model, ko te max_tokens te nui ka pupuri tētahi tono i tō toenga.
  • Hei tepe mō te roa o te whakautu, ka whakamahia te max_tokens e ngā model open-weight hosted, shannon-1.6-lite, shannon-1.6-pro me te shannon-coder-1.
  • I runga i ngā model open-weight hosted ka whakawhāiti i te kuputuhi whakautu. Kāore e tatau ki a ia te reasoning i mua i te whakautu, ā, ka mahi tētahi uara iti iho i te 256 hei 256.
  • Ka mutu tētahi stream o te shannon-1.6-lite, shannon-1.6-pro, shannon-coder-1 rānei, o te whānau Shannon 3 rānei me te finish_reason: "length" ina tapahia te whakautu i tōna tepe.

Ngā tepe me te toenga

Te āhua o te hanga o tētahi utu

  • Ko ngā utu he USD mō ia 1,000,000 token. Ka tatauria tō toenga ki ngā token e uara ana $5.00 mō ia 1,000,000, nō reira ka kī te utu o tētahi model te nui o te toenga ka tango āna token.
  • Kotahi te utu o ngā model Shannon. Ka utua tētahi karanga ki te usage.total_tokens ki taua utu.
  • E toru ngā utu o ngā model open-weight hosted. Ka utua te input i pānuitia mai i te prompt cache ki te utu cached, he 25% o te utu input kua whakaarotia ki te $0.001. Ka utua te toenga o te input ki te utu input, te output ki te utu output.
  • Ka tatau ngā model open-weight hosted i ngā token ki te tokenizer ake a ia model. Ka tohu te GET /v1/models i tēnei hei tokenizer: "original".
  • Ka tatauria te shannon-coder-1 ki ngā karanga, kaua ki ngā token, i runga i te /v1/chat/completions me te /v1/messages. Ka pā tōna utu kua rārangihia i runga i te /v1/responses.

Kei Ngā tepe me te toenga te mea e pupuri ai tētahi tono, ngā tauira mahi o tētahi utu me te whakaaetanga karanga Shannon Coder. Ngā tepe me te toenga

Ngā id model

Tukuna tētahi o ēnei id hei model.

Ngā id 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
  • Kāore e pā te pū nui, pū iti rānei, ā, ka tangohia ngā mokowā huri noa i te id: he model ōrite te SHANNON-3 me te shannon-3.
  • Ka kawe te whakautu i te id kua whakaputaina i te model i ngā wā katoa, ahakoa te tuhinga i tukuna e koe.
  • Ehara tētahi ingoa whakaatu pēnei i te Shannon 3 i te id.

E whakaaetia ana hoki e whā ngā tuhinga poto ake.

Ka tukuna e koe Te model e whakautu ana
shannon-1.6 shannon-1.6-lite
shannon-3-lite shannon-3
shannon-3.1-lite shannon-3.1
shannon-coder shannon-coder-1

Ka whakautua tētahi id kāore i whakaputaina ki te tūnga 400, te momo invalid_request_error me te karere unknown model: <id> i runga i ngā endpoint e toru. Kāore tētahi atu model e whakautu hei whakakapi.

{
  "error": {
    "type": "invalid_request_error",
    "message": "unknown model: gpt-4o"
  }
}
  • Tukuna te model i ngā wā katoa. Ka whakautu te /v1/responses i tētahi tono kāore he mea ki te 400. I runga i te /v1/chat/completions me te /v1/messages ka whakautua taua tono e te shannon-1.6-lite.
  • Ka whakaae te POST /v1/tokenize me te POST /v1/messages/count_tokens ki ngā id open-weight hosted.

Rārangi i ngā model

Ka whakahoki te GET /v1/models i ngā model o tēnei whārangi hei JSON, me ngā matapihi horopaki, ngā utu me ngā āheinga ōrite. Whakamahia hei whakakī i tētahi kaitīpako model, hei tirotiro rānei i tētahi utu i roto i te waehere.

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)

Te whakautu, kua whakapotoa ki tētahi urunga mō ia momo:

200 JSON
{
  "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"
      ]
    }
  ]
}
Āpure Momo Whakaahuatanga Kei runga i
object string Ko list i ngā wā katoa.
data array Kotahi urunga mō ia model ka taea te karanga.
data[].id string Te id ka tukuna e koe hei model. Ia model
data[].object string Ko model i ngā wā katoa. Ia model
data[].created integer He tau pūmau, he ōrite mō ia model: 1700000000. Ia model
data[].owned_by string Ko shannon-ai i ngā wā katoa. Ia model
data[].display_name string He ingoa whakaatu. Kāore e whakaaetia hei model. Ia model
data[].context_window integer Te matapihi horopaki i ngā token. Ia model
data[].max_output_tokens integer Te tahua output nui rawa ka taea e tētahi tono te tono: 65536 i ia model. Ia model
data[].pricing.input_per_million_usd number Te utu mō te 1,000,000 token input i te USD. Ia model
data[].pricing.cached_input_per_million_usd number Te utu mō te 1,000,000 token input i pānuitia mai i te prompt cache. Ngā model open-weight hosted
data[].pricing.output_per_million_usd number Te utu mō te 1,000,000 token output i te USD. I runga i tētahi model Shannon he ōrite ki te utu input. Ia model
data[].pricing.currency string Ko USD i ngā wā katoa. Ia model
data[].tokenizer string original: ka tatauria ngā token ki te tokenizer ake a te model. Ngā model open-weight hosted
data[].capabilities.tools boolean Ka taea e te model te karanga i ngā tool ka whakapuakina e koe. true i ia model. Ia model
data[].capabilities.vision boolean Ka pānui te model i ngā whakaahua. Ia model
data[].capabilities.reasoning boolean Ka whakaaro te model i mua i tana whakautu. Ia model
data[].capabilities.reasoning_effort boolean Ka whai pānga te āpure tono reasoning_effort ki tēnei model. Ia model
data[].capabilities.streaming boolean true i ia model. Ia model
data[].capabilities.json_schema boolean Ka whai pānga te response_format o te momo json_schema. Ia model
data[].capabilities.response_format boolean Ka whai pānga te response_format, tae noa ki te iti rawa hei ahanoa JSON. Ngā model open-weight hosted
data[].capabilities.prompt_caching boolean Ka cached ngā tīmatanga prompt e hoki mai ana, ā, ka utua ki te utu cached. Ngā model open-weight hosted
data[].endpoints array Ngā ara e whakaae ana ki tēnei id. Ia model
  • Kāore he key e hiahiatia e te GET /v1/models. He ōrite te whakautu mō ia kaikaranga, ā, kāore e whakawhirinaki ki tō mahere.
  • Kei te rārangi ngā id ka taea te karanga. Ka waihohia i reira tētahi model kāore e taea te ratoa i tēnei wā.
  • Kāore he endpoint mō tētahi model kotahi: ka whakautu te GET /v1/models/{id} ki te 404 me te momo not_found_error.