Skip to main content
Gemini 4 Carbon supports text generation and image, audio, and video understanding through Chat Completions and Responses. Reasoning defaults to medium. Use low, medium, or high; minimal and thinking budgets are not supported. Output defaults to 32,768 tokens synchronously and 65,536 with streaming. Set a larger budget for automatic continuation within the shared 1,048,576-token input and output window. Frequency and presence penalties and multiple candidates are not supported. End conversation history with a user message or tool results; return tool results in call order. Custom temperature, top_p, and top_k values are ignored. Implicit prompt caching is supported.

Model ID

Pass this exact public ID in the top-level model field.

Supported modes

  • chat_completions
  • responses_api
  • visual_understanding
  • text_generation

Inputs

  • Text
  • Image
  • Video
  • Audio

Outputs

  • Text

Capabilities

  • Not specified

Additional capabilities

  • Chat completions
  • Function calling
  • Code execution
  • Computer use
  • Responses API
  • Video understanding
  • SSE streaming
  • Multimodal understanding
  • Reasoning
  • Structured output See Function calling for tool declarations and result submission.
See Gemini tools for built-in tools, conversation replay, and streaming requirements.

Technical details

Availability and final customer prices can change. Check the Dashboard model catalog before production use.
See the API reference overview for authentication, request conventions, and response schemas.