LLM Models API¶
Note
This feature is available with the Robusta SaaS platform and self-hosted commercial plans. It is not available in the open-source version.
Use this endpoint to list the Robusta-hosted LLM models available to your account, along with the account's default model. This is useful if you are building a tool that lets a user pick a model, or that needs to know which model Holmes will use by default.
POST https://api.robusta.dev/api/llm/models¶
Request Body¶
Parameter |
Type |
Description |
Required |
|---|---|---|---|
|
string |
The unique account identifier (found in the Robusta UI under Settings -> General). |
Yes |
|
string |
Unused for API key authentication; pass an empty string. |
Yes |
Authentication¶
Send your Robusta API key as a Bearer token:
Authorization: Bearer <API_KEY>
The key must have the "Robusta AI" resource with "Write" permission. You can generate this token in the platform by navigating to Settings -> API Keys -> New API Key, and creating a key with the "Robusta AI" resource and "Write" permission.
Example Request¶
curl -X POST 'https://api.robusta.dev/api/llm/models' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer API-KEY' \
--data '{
"account_id": "ACCOUNT_ID",
"session_token": ""
}'
Example Response¶
{
"models": ["claude-sonnet-4-5", "gpt-5.4"],
"models_holmes_args": {},
"default_model": "claude-sonnet-4-5"
}
Response Fields¶
Field |
Type |
Description |
|---|---|---|
|
array |
Robusta-hosted model names available to the account. |
|
object |
Per-model arguments Holmes needs to invoke each listed model. |
|
string |
The account's fallback Robusta-hosted model — used when a requested model is unavailable. Not necessarily the account's overall default, which may be a model defined on a connected agent instead of a Robusta-hosted one. |
Error Responses¶
Status |
Error |
Description |
|---|---|---|
401 |
Unauthorized |
Invalid or missing API key. |
403 |
Forbidden |
API key lacks required "Robusta AI: Write" permission. |
503 |
Service Unavailable |
LLM registry is not configured, or account model settings are unavailable. |
Other Versions¶
/api/llm/models/v2 and /api/llm/models/v3 use the same authentication and request body, and return a more detailed per-model payload (a model-name-keyed object rather than a flat list, plus the account's model preferences on v3). They power the Robusta Platform's own model-picker UI; most integrations should use /api/llm/models above.
See Also¶
Holmes Chat API — the
modelfield on a chat request accepts any model name returned here.