createImagesModels()/ImagesProvider/createImagesProvider() give image
generation the same shape as chat: sync model reads, explicit async
refresh(provider?) with in-flight dedupe, provider-resolved auth, and
never-rejecting generateImages() (failures return AssistantImages with
stopReason error). Auth resolution is shared with the chat side via the
free-standing resolveProviderAuth() in auth/resolve.ts, which also owns
ModelsError; both collections pass their store/context as arguments.
The OpenRouter implementation moves to api/openrouter-images.ts with a
lazy wrapper; openrouterImagesProvider() factory plus
builtinImagesProviders()/builtinImagesModels() land in providers/all.
The ImagesProvider id type alias is renamed to ImagesProviderId
(mirror of Provider -> ProviderId). The old global image API
(getImageModel*, generateImages, registerImagesApiProvider) stays on
/compat, its registration shim repointed at the moved implementation.
README: Quick Start uses builtinModels(), the full streaming event
switch, image generation and the development checklist are restored in
full, image generation documents the new collections with compat noted
for the old API, plus the review fixes (builtinModels options,
credential-store mention for browsers, ImagesModels notes).
Provider.getModels() is sync-only (last-known list; must not throw) with
an optional refreshModels() where dynamic providers fetch. The
sync-or-async union invited latent sync assumptions that would detonate
on the first dynamic provider; async-only reads would force sync
consumer surfaces (extension find/getAll) through Promises. Sync reads
plus an explicit refresh verb keeps the contract single and the
staleness visible.
Models.getModels()/getModel() are sync best-effort reads;
Models.refresh(provider?) rejects with ModelsError(model_source) for a
single provider and is concurrent best-effort across all providers.
createProvider() takes a models array plus an optional refreshModels
fetcher (stored on success, in-flight calls deduped, list unchanged on
rejection). forceRefresh options are gone.
Also finishes the in-progress AuthStorage fallbackResolver removal
(drops the now-unused includeFallback option from getApiKey).
When Bedrock rejects a request with "data retention mode '<mode>' is not
available for this model", append a pointer to the AWS data retention
documentation so users can configure a supported mode.
Azure Foundry deploys gpt-5.4 and gpt-5.5 with a 1,050,000 context
window, but the Azure provider cloned OpenAI-direct's 272k API limit.
Override the context window in the Azure derivation.
Also fix gpt-5-pro maxTokens, which upstream metadata set to 272000 (a
duplicate of the input sub-limit) instead of the actual 128000 max
output; corrected at the source so OpenAI-direct and Azure both match.
closes#5559
- Update GitHub Copilot tests to use available models (claude-haiku-4.5, gemini-2.5-pro)
- Update OpenRouter tests to use available models (google/gemini-2.5-flash, deepseek/deepseek-chat)
- Fix over-broad sed that incorrectly renamed OpenAI Responses gpt-4o test
- Remove nvidia.md from repo root
- Update packages/ai/CHANGELOG.md
Armin Ronacher