LLM pricing index.
Frontier and open-weight model pricing - searchable, sortable. Sourced from LiteLLM, which tracks provider pricing as soon as it's published.
Notes
- Source: the LiteLLM model_prices_and_context_window registry - a community-maintained ledger of frontier model pricing, context windows, and capabilities.
- Units: all prices are USD per 1 million tokens. Cache prices, where listed by the provider, apply only to prompt-cache reads/writes - see the provider's pricing page for cache TTL and eligibility.
- Capabilities: reasoning, vision, audio, tool use, and prompt caching are derived from LiteLLM's flags. Anything not flagged isn't shown.
- Refresh cadence: once a day.
[ - / CTA ]
Spending more
than you should?
Let's find where your cloud and AI spend can work harder.