gk ai models
List the GitKraken-hosted AI models available for the model override
Synopsis
List the models the GitKraken AI backend offers, including which is the default
and each model’s token limits. Use a model id with gk ai provider set gitkraken --model <id>
to override the model used for all GitKraken-hosted AI calls.
The list is task-scoped on the backend (models are registered per task type); this shows the models for the default task type, which is the set valid for the GitKraken model override. Results are cached for 4 hours; use –refresh to force a fresh fetch.
gk ai models [flags]
Examples
gk ai models
gk ai models --refresh
gk ai models --json
gk ai models --json --fields modelId,modelName
Options
--fields string Comma-separated list of fields to keep in JSON output (e.g. modelId,modelName); requires --json
-h, --help help for models
--refresh Bypass the 4h cache and refetch the model list
Options inherited from parent commands
--json Emit machine-readable JSON on stdout (GK_OUTPUT=json sets it session-wide)
--no-telemetry Disable telemetry (OTel spans, Sentry) for this invocation
--session string (Optional) Isolate auth and cache under a named session
SEE ALSO
- gk ai - Use AI in the CLI