Llm Models Foundation Model Retrieve 2

GET

https://api.respan.ai/llm_models/foundation_model/:model_name/

GET

/llm_models/foundation_model/:model_name/

$ curl https://api.respan.ai/llm_models/foundation_model/model_name/ \
>      -H "Authorization: Bearer <token>" \
>      -H "Content-Type: application/json"

200Retrieved

1 {
2   "id": 42,
3   "variants": "v1.0, v1.1, v2.0",
4   "model_name": "gpt-4-respan",
5   "updated_at": "2024-01-15T09:30:00Z",
6   "display_name": "GPT-4 Respan Edition",
7   "speed": 0.85,
8   "max_context_window": 8192,
9   "model_size": 175000000000,
10   "mmlu_score": 87.5,
11   "mt_bench_score": 92.3,
12   "big_bench_score": 88.7,
13   "input_cost": 0.0004,
14   "output_cost": 0.0006,
15   "rate_limit": 1000,
16   "token_rate_limit": 50000,
17   "multilingual": 1,
18   "streaming_support": 1,
19   "function_call": 1,
20   "enforce_function_call": 0,
21   "weight": 1,
22   "image_support": 0,
23   "hf_url": "https://huggingface.co/openai/gpt-4-respan",
24   "model_description": "GPT-4 variant optimized for Respan AI observability and evaluation platform, supporting large context windows and high throughput.",
25   "model_params": [
26     "temperature=0.7",
27     "top_p=0.9",
28     "max_tokens=2048"
29   ],
30   "total_requests": 125000,
31   "total_cost": 3500.75,
32   "total_tokens": 450000000,
33   "total_completion_tokens": 300000000,
34   "total_prompt_tokens": 150000000,
35   "avg_tps": 120.5
36 }

Foundation model detail by model_name. Auth optional — API key OR JWT parsed if present, anonymous allowed. Serializer filters variants by org when authenticated. See FoundationModelView for why the optional mixin replaces the bare JWT authenticator (it 401’d valid API-key callers).

Authentication

AuthorizationBearer

JWT access token or Respan API key

Path parameters

model_namestringRequired

Response

idintegerRead-only

variantsstringRead-only

model_namestring<=255 characters

updated_atdatetimeRead-only

display_namestringOptional<=255 characters

speeddoubleOptional

max_context_windowintegerOptional-2147483648-2147483647

model_sizeintegerOptional-2147483648-2147483647

mmlu_scoredoubleOptional

mt_bench_scoredoubleOptional

big_bench_scoredoubleOptional

input_costdoubleOptional

output_costdoubleOptional

rate_limitintegerOptional-2147483648-2147483647

token_rate_limitintegerOptional-2147483648-2147483647

multilingualintegerOptional-2147483648-2147483647

streaming_supportintegerOptional-2147483648-2147483647

function_callintegerOptional-2147483648-2147483647

enforce_function_callintegerOptional-2147483648-2147483647

weightdoubleOptional

image_supportintegerOptional-2147483648-2147483647

hf_urlstringOptionalformat: "uri"<=200 characters

model_descriptionstringOptional

model_paramslist of stringsOptional

total_requestsintegerOptional-2147483648-2147483647

total_costdoubleOptional

total_tokenslongOptional-9223372036854776000-9223372036854776000

total_completion_tokenslongOptional-9223372036854776000-9223372036854776000

total_prompt_tokenslongOptional-9223372036854776000-9223372036854776000

avg_tpsdoubleOptional