fix: replace retired Gemini model pins with -latest aliases - #402
Open
dakshcodez wants to merge 1 commit into
Open
fix: replace retired Gemini model pins with -latest aliases#402dakshcodez wants to merge 1 commit into
dakshcodez wants to merge 1 commit into
Conversation
gemini-2.0-flash, gemini-2.0-flash-lite, gemini-2.5-flash, gemini-2.5-flash-lite, and gemini-2.5-pro all fail on the OpenAI-compat endpoint for new API keys — the 2.5 models 404 with 'no longer available to new users', and the 2.0 models are quota-exhausted with nothing to fall back to. Switch to gemini-flash-latest, gemini-flash-lite-latest, and gemini-pro-latest, which resolve to whatever Google currently serves under each tier instead of a dated snapshot, so this doesn't recur every time a pinned version is retired. Update .env.example and README references to match. Verified against the live API and via score.py end to end: section extraction completes with no 404s. Fixes interviewstreet#378
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Fixes #378.
All five Gemini models shipped in
providers.jsonfail on a fresh API key:gemini-2.0-flash/gemini-2.0-flash-litegemini-2.5-flash/gemini-2.5-flash-lite/gemini-2.5-proThe 404s are unambiguous: these model versions are gone at the model level, independent of quota/billing.
.env.exampleandREADME.mdboth point new contributors atgemini-2.5-pro, so this is the first thing a fresh clone hits.-latestaliases:gemini-flash-latest,gemini-flash-lite-latest,gemini-pro-latest. These resolve to whatever Google currently serves per tier instead of a specific retired snapshot, so this doesn't recur every time a version is deprecated (the same class of break already happened once going from 2.0 to 2.5)..env.exampleand the twoREADME.mdreferences (env var table + Gemini setup section) to match.Verification
Confirmed directly against
https://generativelanguage.googleapis.com/v1beta/openai/chat/completionswith a live key:gemini-2.5-pro,gemini-2.5-flash→ 404 "no longer available to new users"gemini-flash-latest,gemini-flash-lite-latest→ 200gemini-pro-latest→ valid (429 quota on this key at test time, not 404)Ran the CLI end to end:
with
DEFAULT_MODEL=gemini-flash-latest— basics-section extraction completed with no 404s or retries (previously failed immediately with the abort seen in #378).Test plan
score.pyend to end against a real resume withgemini-flash-latest