GPT-5.4 Pro
Key: gpt-5.4-pro
Provider: OpenAI
Support: runnable
Capabilities
chat · reasoning · tools · vision · json_schema
Specifications
| Field | Value |
|---|---|
| Context window | 400 000 tokens |
| Input modalities | text · image · document |
| Output modalities | text |
When to use
The most capable model in the GPT-5.4 tier. Best for complex multi-step tasks, long agentic workflows that span many tool calls, and vision-heavy lens designs that require both image understanding and structured output.