You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Refreshed generated provider catalogs. OpenRouter pricing and output limits follow its model API, including Nemotron 3 Ultra's 32,768-token output limit and DeepSeek Flash Latest's 943,718-token output limit. See the release verification for the exact source checks.
Removed
Removed fireworks/accounts/fireworks/models/mistral-large-3-fp8 from the built-in serverless catalog, following models.dev #7286. Fireworks still offers the model through on-demand deployments, but does not support serverless access. Existing selections of this built-in ID must be changed; on-demand deployments require their own endpoint configuration.
Fixed
Case.dev model image support now comes only from modalities.input. Models without that field are text-only; vision and multimodal tags no longer grant image input (CD-1621, #71).