LLM model pricing compared
Input and output pricing, context window, and tokenizer for all 15 OpenAI, Anthropic, and Google models, cheapest first. Rates are per million tokens and reflect promotional pricing where it applies.
| Model | Input | Output | Context |
|---|---|---|---|
GPT-4o miniOpenAI | $0.15 | $0.60 | 128K |
GPT-5.6 LunaOpenAI | $0.20 | $1.20 | 1,050K |
Gemini 3.1 Flash-LiteGoogle | $0.25 | $1.50 | 1,000K |
Gemini 2.5 FlashGoogle | $0.30 | $2.50 | 1,000K |
GPT-3.5 TurboOpenAI | $0.50 | $1.50 | 16.385K |
Claude Haiku 4.5Anthropic | $1.00 | $5.00 | 200K |
Gemini 3.6 FlashGoogle | $1.50 | $7.50 | 1,000K |
GPT-5.6 TerraOpenAI | $2.00 | $12.00 | 1,050K |
GPT-4.1OpenAI | $2.00 | $8.00 | 1,047.576K |
Claude Sonnet 5Anthropic | $2.00 | $10.00 | 1,000K |
Gemini 3.1 ProGoogle | $2.00 | $12.00 | 1,000K |
GPT-4oOpenAI | $2.50 | $10.00 | 128K |
GPT-5.6 SolOpenAI | $5.00 | $30.00 | 1,050K |
Claude Opus 5Anthropic | $5.00 | $25.00 | 1,000K |
GPT-4 TurboOpenAI | $10.00 | $30.00 | 128K |
Per-million rates are a starting point, not a bill. Several models charge more for the entire request once it crosses a long-context threshold — each model’s own page states whether that applies and at what point.