List Models
GET /v1/models — list all models callable by the current token.
List Models
Returns all models callable by the current token:
curl https://api.tokenhub.belivestudios.com/v1/models \
-H "Authorization: Bearer sk-your-key"Example response:
{
"object": "list",
"data": [
{
"id": "gpt-5.6-sol",
"object": "model",
"created": 1626777600,
"owned_by": "openai",
"supported_endpoint_types": [
"openai"
]
},
{
"id": "claude-fable-5",
"object": "model",
"created": 1626777600,
"owned_by": "claude",
"supported_endpoint_types": [
"anthropic",
"openai"
]
}
]
}Note
The list respects token permissions: if a token is restricted to specific models, only that subset is returned.
Recommendations
Fetch the model list at startup: don't hard-code model IDs — validate against /v1/models first. Misspelled model names are the most common cause of 400 errors.
Gemini thinking-mode compatibility
/v1/models discovers the models callable by the current token and does not add thinking-mode fields to the response. The following behavior was verified through the BeLive Studio Google-native endpoint on September 10, 2026:
| Model ID | Can thinking be disabled? | Parameter |
|---|---|---|
gemini-3.5-flash | Yes | generationConfig.thinkingConfig.thinkingBudget: 0 |
gemini-3.6-flash | Yes | generationConfig.thinkingConfig.thinkingBudget: 0 |
gemini-3.1-pro-preview | Yes | generationConfig.thinkingConfig.thinkingBudget: 0 |
gemini-3.7-flash | Not guaranteed | No reliable off switch; thinkingLevel: "LOW" only reduces the level |
See Gemini (Google-Compatible) for a complete request example, verification steps, and the gemini-3.7-flash caveat.
For quota queries see Tokens & Quota.