Checked on 2026-10-14 against ai.google.dev/gemini-api/docs/models, as part of assessing yoyo's #1020 (bedrock/google output-ceiling warnings).
Findings:
- gemini-2.5-flash, yoyo's default google model (src/providers.rs default_model_for_provider): the Gemini API now limits 2.5 access to users who have actively used it before. Google says new projects should use 3.5 Flash-Lite or 3.8 Flash. On Vertex/Gemini Enterprise, gemini-2.5-flash retires 2026-10-20.
- gemini-2.5-flash output token limit is 65,536 (input 1,048,576). yoyo's google arm warns at 8192 as a 'declared output ceiling', which is false.
- Current Gemini API ids include gemini-4-argon (flagship, 1M context, decodes up to 1M), gemini-3.8-flash (output 65,536, thinking low/medium/high, minimal returns an error), gemini-3.6-flash, gemini-3.5-flash-lite, gemini-3.1-flash-lite, gemini-3.1-pro-preview, gemini-3-flash-preview.
- yoyo's known-model lists (src/providers.rs, src/commands.rs) carry gemini-3.0-pro and gemini-3.0-flash. Those ids do not appear in Google's list (the real 3.x ids are the ones above).
- Separately, Bedrock: yoyo warns at 16000 for anthropic.claude-sonnet-4-6, which is yoyo's request default, not the model's maximum.
Sources: https://ai.google.dev/gemini-api/docs/models ; https://ai.google.dev/gemini-api/docs/models/gemini-2.5-flash ; https://ai.google.dev/gemini-api/docs/models/gemini-3.8-flash ; https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/gemini/2-5-flash