LLM API pricing · 2026
Current per-1K token rates for OpenAI, Anthropic, Google, Mistral, DeepSeek, and xAI. Click any model for a dedicated cost calculator + cheaper-alternative comparison.
| Model | Input / 1K | Output / 1K | |||
|---|---|---|---|---|---|
| OpenAI | |||||
| GPT-4o gpt-4o | $0.00250 | $0.01000 | Calculator → | ||
| GPT-4o mini gpt-4o-mini | $0.00015 | $0.00060 | Calculator → | ||
| GPT-4 Turbo gpt-4-turbo | $0.01000 | $0.03000 | Calculator → | ||
| GPT-4 gpt-4 | $0.03000 | $0.06000 | Calculator → | ||
| o1 o1 | $0.01500 | $0.06000 | Calculator → | ||
| o1-mini o1-mini | $0.00300 | $0.01200 | Calculator → | ||
| Anthropic | |||||
| Claude Sonnet 4 claude-sonnet-4 | $0.00300 | $0.01500 | Calculator → | ||
| Claude 3.5 Sonnet claude-3-5-sonnet | $0.00300 | $0.01500 | Calculator → | ||
| Claude 3.5 Haiku claude-3-5-haiku | $0.00080 | $0.00400 | Calculator → | ||
| Claude 3 Opus claude-3-opus | $0.01500 | $0.07500 | Calculator → | ||
| Gemini 2.5 Pro gemini-2-5-pro | $0.00125 | $0.01000 | Calculator → | ||
| Gemini 2.5 Flash gemini-2-5-flash | $0.00015 | $0.00060 | Calculator → | ||
| Mistral | |||||
| Mistral Large mistral-large | $0.00200 | $0.00600 | Calculator → | ||
| DeepSeek | |||||
| DeepSeek Chat deepseek-chat | $0.00014 | $0.00028 | Calculator → | ||
| DeepSeek Reasoner deepseek-reasoner | $0.00055 | $0.00219 | Calculator → | ||
| xAI | |||||
| Grok grok-beta | $0.00500 | $0.01500 | Calculator → | ||
Prices sourced from each provider's public API rate card. Last reviewed April 2026. 16 models tracked.
Ranked by input token price. The cheapest flagships for high-volume workloads.
DeepSeek Chat
Ridiculously cheap OpenAI-compatible model. Great for high-volume workloads.
GPT-4o mini
The cheapest credible OpenAI model. 80% of GPT-4o quality at 1/17th the input cost.
Gemini 2.5 Flash
Google's GPT-4o-mini equivalent. Cheap + huge context + fast.
DeepSeek Reasoner
DeepSeek's reasoning model. Cheap alternative to o1-mini for structured thinking.
Claude 3.5 Haiku
Anthropic's budget pick. Surprisingly capable for the price.
Gemini 2.5 Pro
Google's frontier model. 2M-token context is the killer feature.
Next step
Upload your actual usage (OpenAI / Anthropic / Google export, or paste numbers). We show the exact dollar amount TrackCost's cache + model routing would have cut from last month's bill.