feat(providers): refresh curated model catalogs and retire Imagen 4 by backryun · Pull Request #10537 · diegosouzapw/OmniRoute

backryun · 2026-08-16T12:39:34Z

Summary

Refresh the curated model catalogs for Gemini-family providers, Perplexity Web, PromptQL, Notion Web, KIE, TinyCMS, Conol, OpenCode Zen, and LMArena.
Preserve provider-specific upstream identifiers instead of forcing one global Gemini Flash name.
Register current Gemini 3.7 promotional pricing and the published CheaperInference and OrcaRouter rates while keeping Gemini Web on its existing flat-rate path.
Restore the stable notion-ai default in the Notion fallback catalog and synchronize the captured PromptQL model references and vision flags.
Retire Google Imagen 4 from the direct Gemini image path and KIE catalog, including its registry, executor, rate-limit metadata, documentation, and obsolete tests.

Catalog updates

Gemini Flash and pricing

Refresh Antigravity/AGY, Gemini, Gemini Web, GitHub Copilot, CheaperInference, Devin, OrcaRouter, LMArena, Conol, Notion, PromptQL, KIE, and TinyCMS entries according to each provider's currently exposed identifiers.
Remove retired Gemini 3.5/3.6 Flash entries only where the provider has a verified replacement; providers that still expose an older identifier remain provider-scoped.
Keep OrcaRouter on google/gemini-3.6-flash and use its published provider-specific rates.
Remove the stale Gemini 3.5 free-catalog entry and obsolete provider-specific pricing coverage.

Web and curated provider catalogs

Replace the PromptQL fallback snapshot with the captured model_reference, model_id, config IDs, and vision capability flags.
Refresh Notion's friendly IDs and internal food codenames while retaining notion-ai as the stable default.
Refresh KIE, TinyCMS, and Conol to their current curated model sets.
Refresh OpenCode Zen newest-first, remove delisted models, and align the context-overflow fixture with a live free-tier model.
Refresh Perplexity Web mappings and update the LMArena Gemini 3.6 direct-chat mapping.

Google Imagen 4 retirement

Remove:
- imagen-4.0-generate-001
- imagen-4.0-ultra-generate-001
- imagen-4.0-fast-generate-001
Delete the Gemini Imagen image-provider registry and dedicated :predict executor.
Remove Imagen 4 rate-limit rows, KIE aliases, discovery/parser behavior, README advertising, and obsolete tests.
Keep unrelated imagen3 providers and KIE Nano Banana models unchanged.

Why

Several static catalogs had drifted from the model identifiers currently exposed by their upstream providers. The drift advertised retired models, hid newer curated entries, and left dead Google Imagen 4 execution paths in the image registry. This refresh makes each provider catalog match its own upstream contract without reintroducing intentionally removed models.

Validation

Focused Node unit suite: 328 passed, 0 failed
npm run test:vitest: 40 files, 362 tests passed
npm run lint: passed
npm run typecheck:core: passed
node scripts/check/check-docs-sync.mjs: passed
npm run check:any-budget:t11: passed
node scripts/check/check-tracked-artifacts.mjs: passed
git diff --check: passed

Prettier was intentionally not run for this catalog refresh.

diegosouzapw · 2026-08-17T05:36:25Z

The refresh is useful but not merge-ready: retirement filters remove IDs still present in model specs and quota/output-cap contracts, while the pricing test was deleted without replacement assertions. Please reconcile cross-layer contracts, restore pricing coverage, and rerun the relevant gates.

backryun · 2026-08-17T05:53:21Z

The refresh is useful but not merge-ready: retirement filters remove IDs still present in model specs and quota/output-cap contracts, while the pricing test was deleted without replacement assertions. Please reconcile cross-layer contracts, restore pricing coverage, and rerun the relevant gates.

working on it right now.

backryun · 2026-08-17T15:02:42Z

@diegosouzapw Finished, I hope you satisfy it.

…10749) * fix(ci): route the Gemini Web b64_json download error through sanitizeErrorMessage() open-sse/handlers/imageGeneration/providers/geminiWeb.ts embedded a raw err.message in the b64_json download-failure response, tripping check:error-helper (Hard Rule #12) on release/v3.8.50. Refs #9985. * fix(tests): drain test-drift base-reds left by #10603/#10537 and a stale qwen-web catalog id Several base-reds on release/v3.8.50 (#9985) share one root cause: a legitimate product change landed without updating the test asserting the old behavior. - tests/unit/glm-provider-model-import-route.test.ts (12 tests) and tests/unit/model-sync-route.test.ts (2 tests) predate #10603, which made upstream model sync opt-in (isAutoFetchModelsEnabled() now requires providerSpecificData.autoFetchModels === true) and made manual custom-model overrides survive a sync instead of being demoted. Updated both files to opt in / assert the new preserve-manual-overrides behavior, with a comment citing #10603. - tests/unit/antigravity-model-aliases.test.ts predates #10537, which retired the collapsed 'gemini-3.7-flash' alias (upstream 'gemini-3.7-flash-tiered') in favor of the three directly-callable tiered ids. Dropped the retired id from EXPECTED_FLASH_TIERS. - open-sse/config/freeModelCatalog.data.ts: the qwen-web free-catalog entry still listed the retired 'qwen3.8-max-preview' id instead of the current 'qwen3.8-max' (open-sse/config/providers/registry/qwen/web/index.ts and the executor's compat alias both confirm 'qwen3.8-max' is canonical). Real data drift, not test drift. - src/i18n/messages/zh-TW.json: providers.autoFetchModelsTooltip (added by #10603) used the mainland term 緩存 instead of the zh-TW glossary-canonical 快取, tripping the i18n-glossary-consistency-check base-red. - src/lib/oauth/providers/zed-hosted.ts: removed an unused default export (the named export already covers every consumer) — shaves one symbol off the check:dead-code ratchet (419 -> 418; baseline 415, 3 still outstanding). Refs #9985. --------- Co-authored-by: Markus Hartung <mail@hartmark.se>

…iegosouzapw#10537) * feat(providers): refresh Gemini Flash catalogs and pricing * fix(providers): refresh gemini-web Flash catalog * chore(providers): eliminate Gemini 3.5/3.6 Flash models * feat(providers): refresh Perplexity Web model mappings * feat(providers): refresh PromptQL and Notion catalogs * feat(providers): refresh KIE TinyCMS and Conol catalogs * feat(providers): refresh OpenCode Zen catalog * chore(providers): finish Gemini Flash cleanup * chore(providers): retire Google Imagen 4

…apw#9985) (diegosouzapw#10749) * fix(ci): route the Gemini Web b64_json download error through sanitizeErrorMessage() open-sse/handlers/imageGeneration/providers/geminiWeb.ts embedded a raw err.message in the b64_json download-failure response, tripping check:error-helper (Hard Rule #12) on release/v3.8.50. Refs diegosouzapw#9985. * fix(tests): drain test-drift base-reds left by diegosouzapw#10603/diegosouzapw#10537 and a stale qwen-web catalog id Several base-reds on release/v3.8.50 (diegosouzapw#9985) share one root cause: a legitimate product change landed without updating the test asserting the old behavior. - tests/unit/glm-provider-model-import-route.test.ts (12 tests) and tests/unit/model-sync-route.test.ts (2 tests) predate diegosouzapw#10603, which made upstream model sync opt-in (isAutoFetchModelsEnabled() now requires providerSpecificData.autoFetchModels === true) and made manual custom-model overrides survive a sync instead of being demoted. Updated both files to opt in / assert the new preserve-manual-overrides behavior, with a comment citing diegosouzapw#10603. - tests/unit/antigravity-model-aliases.test.ts predates diegosouzapw#10537, which retired the collapsed 'gemini-3.7-flash' alias (upstream 'gemini-3.7-flash-tiered') in favor of the three directly-callable tiered ids. Dropped the retired id from EXPECTED_FLASH_TIERS. - open-sse/config/freeModelCatalog.data.ts: the qwen-web free-catalog entry still listed the retired 'qwen3.8-max-preview' id instead of the current 'qwen3.8-max' (open-sse/config/providers/registry/qwen/web/index.ts and the executor's compat alias both confirm 'qwen3.8-max' is canonical). Real data drift, not test drift. - src/i18n/messages/zh-TW.json: providers.autoFetchModelsTooltip (added by diegosouzapw#10603) used the mainland term 緩存 instead of the zh-TW glossary-canonical 快取, tripping the i18n-glossary-consistency-check base-red. - src/lib/oauth/providers/zed-hosted.ts: removed an unused default export (the named export already covers every consumer) — shaves one symbol off the check:dead-code ratchet (419 -> 418; baseline 415, 3 still outstanding). Refs diegosouzapw#9985. --------- Co-authored-by: Markus Hartung <mail@hartmark.se>

…iegosouzapw#10537) * feat(providers): refresh Gemini Flash catalogs and pricing * fix(providers): refresh gemini-web Flash catalog * chore(providers): eliminate Gemini 3.5/3.6 Flash models * feat(providers): refresh Perplexity Web model mappings * feat(providers): refresh PromptQL and Notion catalogs * feat(providers): refresh KIE TinyCMS and Conol catalogs * feat(providers): refresh OpenCode Zen catalog * chore(providers): finish Gemini Flash cleanup * chore(providers): retire Google Imagen 4

…apw#9985) (diegosouzapw#10749) * fix(ci): route the Gemini Web b64_json download error through sanitizeErrorMessage() open-sse/handlers/imageGeneration/providers/geminiWeb.ts embedded a raw err.message in the b64_json download-failure response, tripping check:error-helper (Hard Rule diegosouzapw#12) on release/v3.8.50. Refs diegosouzapw#9985. * fix(tests): drain test-drift base-reds left by diegosouzapw#10603/diegosouzapw#10537 and a stale qwen-web catalog id Several base-reds on release/v3.8.50 (diegosouzapw#9985) share one root cause: a legitimate product change landed without updating the test asserting the old behavior. - tests/unit/glm-provider-model-import-route.test.ts (12 tests) and tests/unit/model-sync-route.test.ts (2 tests) predate diegosouzapw#10603, which made upstream model sync opt-in (isAutoFetchModelsEnabled() now requires providerSpecificData.autoFetchModels === true) and made manual custom-model overrides survive a sync instead of being demoted. Updated both files to opt in / assert the new preserve-manual-overrides behavior, with a comment citing diegosouzapw#10603. - tests/unit/antigravity-model-aliases.test.ts predates diegosouzapw#10537, which retired the collapsed 'gemini-3.7-flash' alias (upstream 'gemini-3.7-flash-tiered') in favor of the three directly-callable tiered ids. Dropped the retired id from EXPECTED_FLASH_TIERS. - open-sse/config/freeModelCatalog.data.ts: the qwen-web free-catalog entry still listed the retired 'qwen3.8-max-preview' id instead of the current 'qwen3.8-max' (open-sse/config/providers/registry/qwen/web/index.ts and the executor's compat alias both confirm 'qwen3.8-max' is canonical). Real data drift, not test drift. - src/i18n/messages/zh-TW.json: providers.autoFetchModelsTooltip (added by diegosouzapw#10603) used the mainland term 緩存 instead of the zh-TW glossary-canonical 快取, tripping the i18n-glossary-consistency-check base-red. - src/lib/oauth/providers/zed-hosted.ts: removed an unused default export (the named export already covers every consumer) — shaves one symbol off the check:dead-code ratchet (419 -> 418; baseline 415, 3 still outstanding). Refs diegosouzapw#9985. --------- Co-authored-by: Markus Hartung <mail@hartmark.se>

backryun marked this pull request as ready for review August 16, 2026 12:45

backryun requested a review from diegosouzapw as a code owner August 16, 2026 12:45

backryun force-pushed the agent/refresh-gemini-flash-catalogs branch from e0ec492 to 82504de Compare August 17, 2026 01:54

backryun force-pushed the agent/refresh-gemini-flash-catalogs branch 4 times, most recently from 74011e6 to ce8088b Compare August 17, 2026 11:52

thinh0704hcm mentioned this pull request Aug 17, 2026

cherry-pick(pr-10537): feat(providers): refresh Gemini Flash catalogs and pricing (promo pricing subset) thinh0704hcm/OmniRoute#11

Merged

backryun changed the title ~~feat(providers): refresh Gemini Flash catalogs and pricing~~ feat(providers): refresh curated model catalogs and retire Imagen 4 Aug 17, 2026

backryun force-pushed the agent/refresh-gemini-flash-catalogs branch 6 times, most recently from 8912d3b to b4de9ee Compare August 18, 2026 14:04

backryun mentioned this pull request Aug 18, 2026

feat(providers): publish Gemini 3.7 Flash tiers on the agy provider #10665

Closed

5 tasks done

backryun added 9 commits August 18, 2026 23:43

feat(providers): refresh Gemini Flash catalogs and pricing

4857b80

fix(providers): refresh gemini-web Flash catalog

720054f

chore(providers): eliminate Gemini 3.5/3.6 Flash models

a9ead59

feat(providers): refresh Perplexity Web model mappings

14a9e80

feat(providers): refresh PromptQL and Notion catalogs

6e96266

feat(providers): refresh KIE TinyCMS and Conol catalogs

454e7c1

feat(providers): refresh OpenCode Zen catalog

3029851

chore(providers): finish Gemini Flash cleanup

326b82d

chore(providers): retire Google Imagen 4

bc7c8b8

backryun force-pushed the agent/refresh-gemini-flash-catalogs branch from b4de9ee to bc7c8b8 Compare August 18, 2026 14:43

diegosouzapw merged commit df90591 into diegosouzapw:release/v3.8.50 Aug 18, 2026
4 of 5 checks passed

yawar-aquil mentioned this pull request Aug 18, 2026

feat(providers): add tabitoken gateway and serve hcnsec's four protocols #10668

Merged

11 tasks done

backryun deleted the agent/refresh-gemini-flash-catalogs branch August 19, 2026 06:29

This was referenced Aug 19, 2026

[BUG] Multiple Provider Issues (Gemini/Qwen Web, OpenCode, Antigravity), Windows Binary Mismatch for Embedded Services, and Proxy Pool 429 #10629

Closed

fix(ci): clear remaining base-reds on release/v3.8.50 (refs #9985) #10749

Merged

Navigation Menu

Uh oh!

feat(providers): refresh curated model catalogs and retire Imagen 4 - #10537

feat(providers): refresh curated model catalogs and retire Imagen 4#10537
diegosouzapw merged 9 commits into
diegosouzapw:release/v3.8.50from
backryun:agent/refresh-gemini-flash-catalogs

backryun commented Aug 16, 2026 •

edited

Loading

Uh oh!

diegosouzapw commented Aug 17, 2026

Uh oh!

backryun commented Aug 17, 2026

Uh oh!

backryun commented Aug 17, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

backryun commented Aug 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Catalog updates

Gemini Flash and pricing

Web and curated provider catalogs

Google Imagen 4 retirement

Why

Validation

Uh oh!

diegosouzapw commented Aug 17, 2026

Uh oh!

backryun commented Aug 17, 2026

Uh oh!

backryun commented Aug 17, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

backryun commented Aug 16, 2026 •

edited

Loading