1.1.2

Generation was broken for all Gemini users. Every model the plugin offered had been retired by Google,
so requests to generateContent came back 404 model not found. Indexing still appeared to work,
which made the failure confusing to diagnose — embeddings run on a separate model
(gemini-embedding-001) that is still supported.

What changed

• Model list updated to the current Gemini Flash lineup: gemini-3.7-flash (new default),
  gemini-3.6-flash, and gemini-3.5-flash-lite
• Retired IDs removed: gemini-2.0-flash and gemini-2.0-flash-lite (shut down June 1, 2026) and
  gemini-1.5-pro (shut down September 29, 2025)
• Gemini image OCR updated to the same current model

Action required if you're upgrading

Your saved settings take priority over the new default, so upgrading alone will not fix this. Open
Settings → Deep Notes and select a model from the dropdown, then reload the plugin. Fresh installs
need no action.

No re-indexing is required — your existing vector index is unaffected.

Full Changelog: 1.1.1...1.1.2

Full Changelog: 1.1.1...1.1.2