Sign in
models

Refresh Model Metadata

Changed on

Refresh metadata for models. Supports two modes via discriminated request body:

  • Bulk async: {"source": "all"} triggers background extraction for all local GGUF models (202 Accepted)

  • Single sync: {"source": "model", "repo": "...", "filename": "...", "snapshot": "..."} performs immediate extraction (200 OK)

Requires PowerUser permissions.

post/bodhi/v1/models/refresh

Request

  • Base URL: http://localhost:1135
  • URL: http://localhost:1135/bodhi/v1/models/refresh
  • Auth: OAuth 2 (authorization code, authorization URL /bodhi/v1/auth/initiate, token URL /bodhi/v1/auth/callback, scopes: resource_power_user)

Request body

OR

Example request

{
  "repo": "bartowski/Qwen2.5-3B-Instruct-GGUF",
  "filename": "Qwen2.5-3B-Instruct-Q4_K_M.gguf",
  "snapshot": "8ba1c3c3ee94ba4b86ff92a749ae687dc41fce3f"
}

Response

Metadata refreshed successfully (sync mode)

sourcestring required
aliasstring required
repostring required
filenamestring required
snapshotstring required
accessboolean required

Whether the current principal may run inference on this model.

sizeinteger nullable

Local GGUF file size in bytes (present when the file is resolvable on disk)

Changes