Inference Price Index
Daily benchmark for LLM token pricing
Frontier
Efficient
Open
Loading...
What do these numbers mean?
Each index shows the average cost per 1 million tokens (blended input/output) across models in that tier. Updated daily from OpenAI, Anthropic, Google, and OpenRouter pricing.
Why This Matters
For CTOs
Track pricing deflation to optimize cloud budgets. LLM costs have dropped 90%+ since 2023 - are your contracts keeping pace?
For Developers
Find the cheapest model for your use case. Compare frontier quality vs efficient-tier cost savings in real time.
For Finance
Forecast AI costs with historical trend data. Build accurate projections based on 19+ pricing events across providers.
The Three Tiers
Frontier
Top-tier reasoning models. Best for complex tasks, coding, and analysis. Highest capability, highest cost.
Efficient
Optimized for speed and cost. Great for chat, summarization, and simple tasks. Approximately 6x cheaper than Frontier.
Open
Open-weight models you can self-host. Competitive quality with transparent weights and flexible deployment.
Free API
Get pricing data programmatically:
curl https://inferencepriceindex.com/v1/index/latest
Returns JSON with all indices, model counts, and individual rates. No authentication required.