Gemini Docs

Billing and getting started

• Updated prepaid credit guidance: the minimum prepay / credit purchase is now $5 (was $10) on /gemini-api/docs/billing, /gemini-api/docs/get-started, and /gemini-api/docs/generate-content/get-started.

Image generation

• Documented video-to-image support for Gemini 3.1 Flash Lite Image alongside Gemini 3.1 Flash Image on /gemini-api/docs/image-generation and /gemini-api/docs/generate-content/image-generation.
• Clarified that Nano Banana / image generation can take text, images, video, or a combination, and that video inputs are supported for both 3.1 Flash Image models.
• Extended thinking level controls (minimal / high) docs to cover Gemini 3.1 Flash Lite Image as well as Gemini 3.1 Flash Image.

Model reference

• /gemini-api/docs/models/gemini-3.1-flash-image: supported inputs updated to Text, Image, Video, and PDF.
• /gemini-api/docs/models/gemini-3.1-flash-lite-image: inputs updated to Text, Image, Video, and PDF; capability table refreshed (e.g. function calling not supported; Batch API, flex, and priority inference listed under consumption options).
• /gemini-api/docs/models/gemini-3-pro-image: flex inference and priority inference marked not supported.
• /gemini-api/docs/models/gemini-2.5-flash-image: capability table corrections (e.g. caching and structured outputs not supported).

Video understanding

• On /gemini-api/docs/video-understanding, added guidance to use streaming (stream=True) or background execution (background=True) for long videos or complex prompts to avoid connection/auth timeouts, including under agentic processing.