List available grounded models
const url = 'https://www.cemented.ai/v1/models';const options = {method: 'GET', headers: {Authorization: 'Bearer <token>'}};
try { const response = await fetch(url, options); const data = await response.json(); console.log(data);} catch (error) { console.error(error);}curl --request GET \ --url https://www.cemented.ai/v1/models \ --header 'Authorization: Bearer <token>'Returns the OpenAI-compatible model list plus capability metadata so callers can choose a model programmatically. _maps_to exposes the underlying Anthropic Claude model the grounded pipeline runs on.
Authorizations
Section titled “Authorizations ”Responses
Section titled “ Responses ”List of available models with capability metadata
object
object
Pass this as model on /v1/responses (e.g. “sonnet”).
Underlying Anthropic Claude model used by the pipeline.
Human-readable display name.
Maximum input + output tokens the model accepts in a single request.
Maximum tokens the model will generate in one response.
Whether the model exposes OpenAI-style tool_use events. Currently always false; grounded search runs internally.
Always true — every response is grounded in retrieved sources with inline citations.
Search backends configured for the deployment (e.g. ["exa", "perplexity"]).
Example
{ "object": "list", "data": [ { "id": "sonnet", "object": "model", "created": 0, "owned_by": "anthropic", "_maps_to": "claude-sonnet-5", "label": "Sonnet 5", "context_length": 200000, "max_output_tokens": 128000, "supports_streaming": true, "supports_tools": false, "grounded": true, "search_providers": [ "exa", "perplexity" ] }, { "id": "opus", "object": "model", "created": 0, "owned_by": "anthropic", "_maps_to": "claude-opus-5-5", "label": "Opus 5.5", "context_length": 200000, "max_output_tokens": 128000, "supports_streaming": true, "supports_tools": false, "grounded": true, "search_providers": [ "exa", "perplexity" ] } ]}Unauthorized
object
object
Example generated
{ "error": { "message": "example", "type": "example", "code": "example" }}