benchr
Pricing Index · Updated June 2026

AI Model Pricing Hub

Live API pricing comparison across all major AI models including OpenAI, Anthropic Claude, Google Gemini, DeepSeek, and open-weights. Sourced directly from official docs.

Data from models.json Neutral reference — never sponsored

Global API Pricing Table

Model Provider Input / 1M Output / 1M Cached Input / 1M
Loading pricing data...

Individual Model Pricing Guides

Click any model below to see its detailed pricing guide, hardware requirements, batch discounts, and prompt caching structures.

How to optimize your AI model API costs

Scaling an AI product requires cost optimization. Consider the following strategies:

  1. Implement Prompt Caching: Keep system prompts and RAG contexts static to receive up to a 90% discount on input tokens.
  2. Use Batch APIs: Submit non-real-time workloads (like dataset evaluation) via Batch APIs for a flat 50% discount.
  3. Route tasks dynamically: Send simple queries (routing, simple classification) to small models like DeepSeek V4-Flash or GPT-5 Mini, reserving expensive flagship models like Claude Opus for complex code or reasoning.