feat(models): add gemini-3.1-flash-lite-image to aistudio catalog

[xyonium]

Summary

• add gemini-3.1-flash-lite-image (Nano Banana 2 Lite) to the aistudio section
• metadata mirrors the official Generative Language API v1beta/models response (fetched 2026-09-10): 65536 input/output token limits, generateContent/countTokens/batchGenerateContent methods, text+image input/output modalities
• the thinking field is omitted to match the API response, which reports only a boolean hint rather than budget levels

Why

The aistudio catalog is now stale for image generation: gemini-2.5-flash-image has become a paid tier in Google AI Studio, and the only image model currently offered on the free tier in the AI Studio playground is gemini-3.1-flash-lite-image (Nano Banana 2 Lite). The catalog still lists only gemini-2.5-flash-image, so AI Studio websocket providers cannot route requests to the free-tier image model — the scheduler only allows models present in this catalog, and CLIProxyAPI refreshes from this repo every 3 hours.

Notes

• the non-lite sibling gemini-3.1-flash-image (Nano Banana 2) is intentionally not included: it is not on the AI Studio free tier
• gemini-3.5-transcribe is planned as a separate PR after real-world testing

Sources

• Model metadata: Generative Language API v1beta/models (live response, 2026-09-10)
• Free/paid tier status: observed in the AI Studio playground model picker (free tier shows only Nano Banana 2 Lite; gemini-2.5-flash-image and gemini-3.1-flash-image now show as paid)

Test plan

• models.json parses; no duplicate IDs in any section
• purely additive 26-line diff; entry placed next to its preview sibling (gemini-3.1-flash-lite-preview)

🤖 Generated with Claude Code