Repository navigation
fix(test-page): match hyphenated model catalog ids - #193
Merged
TheSmallHanCat merged 2 commits intoOct 29, 2026
Merged
TheSmallHanCat merged 2 commits into
TheSmallHanCat merged 2 commits into
Conversation
The tester sidebar filtered /v1/models with legacy underscore prefixes (veo_3_1_t2v, veo_3_1_i2v, veo_3_1_r2v) while the catalog exposes hyphenated ids (veo-3.1-*, omni-1.1-flash-*). No video id matched, so every video section disappeared from the sidebar, Omni included. Classify models once through classifyModel(), normalizing underscores to hyphens so legacy ids still work, and reuse it for the sidebar filter, the type tag and getModelMeta(). Omni reference video needs the image uploader, which the old i2v/r2v substring checks missed entirely. Add the Omni and Extend sections plus their locale keys, and seed Omni entries in the fallback catalog.
Keep I2V/R2V reference uploads for upsample aliases and restore the legacy R2V 1080P fallback model. Add a localized source media ID input for Extend models and serialize it using the existing extend:// message contract. Restore Upsample translations and add browser regression coverage for model and language switching. Validation: 37 i18n/browser tests, Ruff, inline JavaScript syntax, locale key parity, and git diff --check.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
The model tester at
/testfiltered the/v1/modelscatalog with legacy underscore prefixes (veo_3_1_t2v,veo_3_1_i2v,veo_3_1_r2v,gemini-3.0-pro-image). The catalog actually exposes hyphenated ids (veo-3.1-*,omni-1.1-flash-*), so no video id matched any category filter and every video section vanished from the sidebar — including Omni 1.1 Flash.The bug does not stop at the sidebar:
getModelMeta()looked fori2v/r2vsubstrings, which never appear in the hyphenated ids either, so the image uploader was hidden for I2V, R2V and Omni — the exact models that need a reference image.Backend is unaffected:
/v1/modelslists 82 models (9omni-1.1-flash-*) and/v1/chat/completionswithomni-1.1-flash-4s-landscapegenerates video normally.Fix
classifyModel()classifier (image/upsample/i2v/r2v/omni/extend/t2v), normalizing_to-so legacy ids keep working.MODEL_CATEGORIES,getModelType()andgetModelMeta()all derive from it; Omni gets the reference-image uploader (0–3 images).Omni 1.1 Flash 视频and视频续写 (Extend)sections with zh-CN/en locale keys.FALLBACK_MODELS.Verification
omni-1.1-flash-4s-landscapeshows the uploader labelled0-3 imagesand a🎬 Generate videobutton.<video>element.pytest tests: 252 passed, 14 skipped.