OpenAI · 400K context · updated 2026-07-28
Input · what you send
$5
per 1M tokens
Output · what it writes
$30
per 1M tokens
ℹ️ Long-context (>272K input) priced at 2× input / 1.5× output.
333,333 tokens
2,000,000 tokens
For context: the cheapest model tracked here is Gemini 3 Flash at $3/1M output tokens, and the most expensive is GPT-5.5 Pro at $180 — a 60× spread.Try your own budget in the calculator →
| Model | Input / 1M | Output / 1M | vs GPT-5.5 (output) |
|---|---|---|---|
| GPT-5.6 Sol | $5 | $30 | same |
| GPT-5.5 Pro | $30 | $180 | 6× pricier |
| Claude Fable 5 | $10 | $50 | 1.7× pricier |
| GPT-5.6 Terra | $2.50 | $15 | 2× cheaper |
| Claude Opus 4.8 | $5 | $25 | 1.2× cheaper |
| Claude Sonnet 5 | $2 | $10 | 3× cheaper |
| Gemini 3.1 Pro | $2 | $12 | 2.5× cheaper |
| Gemini 3.5 Flash | $1.50 | $9 | 3.3× cheaper |
| GPT-5.6 Luna | $1 | $6 | 5× cheaper |
| Claude Haiku 4.5 | $1 | $5 | 6× cheaper |
| Gemini 3 Flash | $0.50 | $3 | 10× cheaper |
GPT-5.5 supports a 400K token context window. That means in a single API call, you can send up to 400K tokens of input — including your prompt, system instructions, and any documents or code you want the model to process.
In practical terms, 400K tokens is approximately:
Compare all models' context windows on the context windows page.