GPT-4o context window: 128K tokens
GPT-4o has a context window of 128,000 tokens and a maximum output of 16,384 tokens per response. The context window caps the total input the model can attend to in one request. The output cap is separate and limits how long a single response can be. Both figures below are verified against OpenAI documentation.
Verified 2026-08-10.
| Model | Context window (tokens) | Max output (tokens) |
|---|---|---|
| GPT-4o | 128,000 | 16,384 |
| GPT-5 | 400,000 | 128,000 |
| GPT-5.6 Sol | 1,050,000 | 128,000 |
What fits in 128,000 tokens
English text averages about 0.75 words per token, so a 128K window holds roughly 96,000 English words of input. That covers a long report, a large batch of emails, or a moderate codebase slice in one request. It does not cover a full book-length corpus or a large repository, which is where the bigger windows in the table matter.
128,000 tokens ≈ 96,000 English words of input
The 16,384-token output cap is the tighter constraint for many jobs. If you ask GPT-4o to rewrite a document that nearly fills the window, the response cannot exceed 16,384 tokens, so long transformations need chunking even when the input fits.
How GPT-4o compares
GPT-5 raises the window to 400,000 tokens, more than three times the GPT-4o figure, and raises max output to 128,000 tokens. GPT-5.6 Sol goes further, with a 1,050,000-token window and the same 128,000-token output cap. If your workload is bounded by either limit on GPT-4o, both successors remove the bound in practice.
Two things to keep in mind before moving up. Input billing scales with tokens sent, so a bigger window raises the ceiling on what one request can cost. And long prompts can degrade retrieval of material in the middle of the window, an effect well documented in long-context evaluations, so a larger window is not automatically better output on the same task.
Cost at the window limit
GPT-4o is priced at $2.50 per million input tokens and $10.00 per million output tokens. A request that fills the full 128K window costs about $0.32 in input tokens before any output. Full OpenAI rates are on the OpenAI pricing page.
More OpenAI model details are on the OpenAI overview. For every current model side by side, see the context window comparison.