Pricing
BenchGecko provides the most comprehensive AI model pricing comparison, covering every model on every provider with daily price tracking.
Features
Arbitrage Finder
The same model costs different amounts on different providers. GPT-4o on OpenAI direct has one price; the same model on Azure, Together AI, or Fireworks has another. The arbitrage finder surfaces the cheapest provider for any model instantly.
Substitute Finder
Two models with nearly identical benchmark scores can differ by 10x in price. The substitute finder shows cheaper alternatives that match your performance requirements. For example, if you need 85%+ on MMLU, it shows every model that qualifies sorted by price.
Free Tier Tracker
Track which providers offer free tiers, how many tokens per day or month, and rate limit details. Useful for prototyping and development.
Price History
Daily price tracking for every model on every provider. See exactly when prices dropped, by how much, and which providers led the reduction. Historical data goes back to the model's launch date.
Pricing Calculator
The Pricing Calculator lets you estimate monthly costs based on:
- Expected input tokens per day
- Expected output tokens per day
- Number of requests
- Model selection (single or multiple)
Pricing Format
All pricing on BenchGecko is normalized to USD per million tokens:
- Input price: Cost per 1M input tokens
- Output price: Cost per 1M output tokens
Batch API pricing (where available) is shown separately. Volume discounts and committed-use pricing are noted but not used for default comparisons.
Price Alerts
The Changelog captures every price change as it happens. Subscribe to get notified when models you use get cheaper.
Related Pages
- All Models
- Compare Tool -- compare pricing alongside benchmarks
- Provider Directory
- API Documentation
- AI Economy Dashboard -- market-level pricing trends