# list_scoring_models (/mcp/list-scoring-models)



Lists the organization's scoring models, newest first, as [`GET /v1/scoring_models`](/reference/list-scoring-models). Read-only.

| Input     | Description      |
| --------- | ---------------- |
| `api_key` | `/mcp/keys` only |

Returns `scoring_models`, each with its `model_id`, `name`, `status` (`training`, `ready` or `failed`), `criteria` (the context it was built with, or empty), `created_at`, `calibrated_at` (or `null` if it isn't calibrated: one created from criteria alone scores with Finic's estimates until it's calibrated), `archived_at` (or `null`; an archived scoring model can't be used until it's unarchived in the Finic Portal), `examples` and `bad_rate` (the examples it was created or first calibrated with, or 0), `scored` (records scored with it) and `outcomes` (those with an outcome sent to [`calibrate`](/mcp/calibrate)). Check it before [`create_scoring_model`](/mcp/create-scoring-model): the user may already have the scoring model they need. For one scoring model's questions, call [`get_scoring_model`](/mcp/get-scoring-model).