my-google-gemini

Models List

• chat-bison-001: A legacy text-only model optimized for chat conversations.
  • input_token_limit: 4096
  • output_token_limit: 1024
  • temperature: 0.25
• text-bison-001: A legacy model that understands text and generates text as an input.
  • input_token_limit: 8196
  • output_token_limit: 1024
  • temperature: 0.7
• embedding-gecko-001: Obtain a distributed representation of a text.
  • input_token_limit: 1024
  • output_token_limit: 1
  • temperature: None
• gemini-1.0-pro: The best model for scaling across a wide range of tasks.
  • input_token_limit: 30720
  • output_token_limit: 2048
  • temperature: 0.9
• gemini-1.0-pro-001: The best model for scaling across a wide range of tasks. This is stable model that supports tuning.
  • input_token_limit: 30720
  • output_token_limit: 2048
  • temperature: 0.9
• gemini-1.0-pro-latest: The best model for scaling across wide range of tasks. This is the latest model.
  • input_token_limit: 30720
  • output_token_limit: 2048
  • temperature: 0.9
• gemini-1.0-pro-vision-latest: The best image understanding model to handle a broad range of applications.
  • input_token_limit: 12288
  • output_token_limit: 4096
  • temperature: 0.4
• gemini-pro: The best model for scaling across a wide range of tasks.
  • input_token_limit: 30720
  • output_token_limit: 2048
  • temperature: 0.9
• gemini-pro-vision: The best image understanding model to handle a broad range of applications.
  • input_token_limit: 12288
  • output_token_limit: 4096
  • temperature: 0.4
• embedding-001: Obtain a distributed representation of a text.
  • input_token_limit: 2048
  • output_token_limit: 1
  • temperature: None
• aqa: Model trained to return answers to questions that are grounded in provided sources, along with estimating answerable probability.
  • input_token_limit: 7168
  • output_token_limit: 1024
  • temperature: 0.2