API Reference

Models

The live catalog, one entry per served model, with context windows, capabilities, and supported parameters.

Endpoint

GET https://inference.pearlresearch.ai/v1/models

This is the OpenAI-compatible model list, extended with per-model metadata. It's the machine-readable companion to Choosing a model, poll it if your application selects models dynamically or wants to feature-detect before sending a request.

Example

curl https://inference.pearlresearch.ai/v1/models \
  -H "Authorization: Bearer $PEARL_API_KEY"

{
  "data": [
    {
      "id": "zai-org/GLM-5.2",
      "hugging_face_id": "zai-org/GLM-5.2",
      "name": "Z.AI: GLM-5.2",
      "created": 1781568000,
      "input_modalities": ["text"],
      "output_modalities": ["text"],
      "quantization": "fp8",
      "context_length": 1000000,
      "max_output_length": 1000000,
      "pricing": {"prompt": "…", "completion": "…", "input_cache_read": "…"},
      "supported_sampling_parameters": ["temperature", "top_p", "top_k",
        "frequency_penalty", "presence_penalty", "stop", "seed", "max_tokens"],
      "supported_features": ["tools", "json_mode", "reasoning"],
      "description": "GLM-5.2 on the Pearl inference network."
    }
  ]
}

Entry fields

Field	Type	Notes
id	string	The value you pass as `model` in requests.
name	string	Human-readable display name (also used in the dashboard).
hugging_face_id	string	The upstream open-weight checkpoint the deployment serves.
created	integer	Unix timestamp of the model's release.
input_modalities	string[]	What the model accepts, `text`, plus `image` on vision models. Check this before sending multimodal content.
output_modalities	string[]	All catalog models output text.
quantization	string	Serving precision (`fp8`, `int8`).
context_length	integer	Maximum tokens (prompt + response) per request.
max_output_length	integer	Ceiling for generated tokens, the effective upper bound for `max_tokens`.
pricing	object	`prompt` / `completion` / `input_cache_read` (the discounted cached-input rate) as USD-per-token strings. Your organization's billed rates are on the dashboard Models page.
supported_sampling_parameters	string[]	The authoritative per-model list of accepted sampling parameters (e.g. whether `min_p` is available).
supported_features	string[]	Capability flags: `tools`, `json_mode`, `reasoning`. The model guide documents verified capability per model.

Treat context_length, max_output_length, and supported_sampling_parameters as the source of truth for request limits, they change when the catalog changes, without a docs release.