OpenAI: GPT-5.4 Pro
GPT-5.4 Pro is a paid model from OpenAI that accepts text, images, and files as input. It supports tool use and reasoning, which makes it suitable for multi-step workflows and agentic tasks. Its context window reaches 1,050,000 tokens, and it can return up to 128,000 tokens per completion, accommodating long documents and extended outputs in a single session. Structured output support is unconfirmed based on available specifications. At $30.00 per million input tokens and $180.00 per million output tokens, this sits at the expensive end of the market, and buyers should weigh that cost carefully. There is currently no independent benchmark coverage to validate performance claims, so capability is unproven relative to peers. Teams that require large-context multimodal processing with tool and reasoning support may find it worth evaluating, but the price warrants caution until third-party benchmarks establish a clearer picture of where it stands.
- Model ID
- openai/gpt-5.4-pro
- Vendor
- openai
- Tokenizer
- GPT
- Input Modalities
- text, image, file
- Output Modalities
- text
- Max Output
- 128,000 tokens
- Tool Calling
- ✓ supported
- Structured Output
- ✓ supported
- Reasoning Mode
- ✓ supported
- Vision
- ✓ accepts images
- Audio
- no
- Moderated
- yes