Skip to content

Latest commit

 

History

History
44 lines (33 loc) · 2.53 KB

File metadata and controls

44 lines (33 loc) · 2.53 KB

Gemini Model Overview for ADK

ADK supports several Gemini models with different capabilities and price points. Choosing the right model involves balancing performance, capabilities, and cost for your specific use case.

Model Capabilities

Model Description Input Types Best For
gemini-2.5-pro Most powerful thinking model with maximum response accuracy Audio, images, video, text Complex coding, reasoning, multimodal understanding
gemini-2.5-flash Best price-performance balance Audio, images, video, text Low latency, high volume tasks that require thinking
gemini-2.0-flash Newest multimodal model with improved capabilities Audio, images, video, text Low latency, enhanced performance, agentic experiences
gemini-2.0-flash-lite Optimized for efficiency and speed Audio, images, video, text Cost efficiency and low latency
gemini-1.5-flash Versatile performance across diverse tasks Audio, images, video, text Fast and versatile performance
gemini-1.5-flash-8b Smaller, faster model Audio, images, video, text High volume and lower intelligence tasks
gemini-1.5-pro Powerful reasoning capabilities Audio, images, video, text Complex reasoning tasks requiring more intelligence

Pricing

Model Input Price Output Price
gemini-2.5-pro $10.00 / 1M tokens $30.00 / 1M tokens
gemini-2.5-flash $3.50 / 1M tokens $10.50 / 1M tokens
gemini-2.0-flash $3.50 / 1M tokens $10.50 / 1M tokens
gemini-2.0-flash-lite $0.70 / 1M tokens $2.10 / 1M tokens
gemini-1.5-flash $2.50 / 1M tokens $7.50 / 1M tokens
gemini-1.5-flash-8b $0.35 / 1M tokens $1.05 / 1M tokens
gemini-1.5-pro $7.00 / 1M tokens $21.00 / 1M tokens

Token Information

  • A token is approximately 4 characters
  • 100 tokens are roughly 60-80 English words
  • Pricing is calculated based on both input tokens (prompts sent to the model) and output tokens (responses generated by the model)

Model Selection Guidelines

  1. For budget-conscious applications: Start with gemini-2.0-flash-lite
  2. For balanced performance and cost: Use gemini-2.0-flash or gemini-2.5-flash
  3. For complex reasoning tasks: Choose gemini-2.5-pro
  4. For production applications: Prefer stable models over experimental/preview versions

Additional Resources

For the most up-to-date information on Gemini models, visit the official Gemini API documentation.