Gemini 2.0 Flash Thinking Mode

gemini-2.0-flash-thinking-exp is supported!

Ref: https://ai.google.dev/gemini-api/docs/thinking-mode

Other improvements:

• Support frequency_penalty, presence_penalty and max_completion_tokens
• Improve response_format handling, support "json_schema"
• Support responseMimeType: "text/x.enum"
• Misc. important fixes

LearnLM - for teaching and learning

learnlm-1.5-pro-experimental model is supported!

Ref:

• https://ai.google.dev/gemini-api/docs/learnlm
• https://blog.google/outreach-initiatives/education/google-learnlm-gemini-generative-ai/

Other improvements:

• Support embeddings and models API endpoints.
• Serve locally with Node, Deno, Bun.
• Deploy to Deno instruction.
• Support audio input.

Gemini 1.5 Flash & Pro 002

Ready for updated models:

• gemini-1.5-pro-002
• gemini-1.5-flash-002
• gemini-1.5-flash-8b-exp-0924

https://developers.googleblog.com/en/updated-production-ready-gemini-models-reduced-15-pro-pricing-increased-rate-limits-and-more/

Gemini 1.5 Flash

Use gemini-1.5-flash-latest (and other models), specifying it in request.

https://developers.googleblog.com/en/gemini-15-pro-and-15-flash-now-available/

Also: response object now contains usage info.

Gemini 1.5 Pro

Use gemini-1.5-pro-latest model!

https://developers.googleblog.com/2024/04/gemini-15-pro-in-public-preview-with-new-features.html