Computer Use Preview
At a glance
- Price per 1M tokens (input)
- $3.00
- Price per 1M tokens (output)
- $12.00
- Context
- 8,192 tokens
- Free access
- yes (see below)
Refreshed daily; data verified September 14, 2026. Published September 14, 2026.
Computer Use Preview by OpenAI is designed for developers and technical teams building applications that require visual understanding and automated interaction with software interfaces. It excels at tasks such as interpreting screenshots, navigating graphical user interfaces, and executing multi-step workflows based on visual input. The model combines image understanding with tool calling — a capability that allows it to invoke external functions or APIs in response to perceived content — enabling step-by-step reasoning over dynamic digital environments. This approach distinguishes it by focusing on grounding AI actions in real-time visual feedback rather than relying solely on textual descriptions.
Specifications & pricing
| Input (per 1M tokens) | $3.00 |
|---|---|
| Output (per 1M tokens) | $12.00 |
| Context window | 8,192 tokens |
| Max output | 1,024 tokens |
| Capabilities | images, tool calling, reasoning |
LiteLLM community dataset (MIT), verified September 14, 2026. Official OpenAI pricing.
What Computer Use Preview would cost on your workload — run it through the cost calculator →
Where to try Computer Use Preview for free
- OpenAI offers a free chat — ChatGPT (free plan). A vendor's free chat may run a different model from the same family — the exact model is not guaranteed.
Frequently asked questions
What types of tasks is Computer Use Preview best suited for?+
It is ideal for automating repetitive UI-based workflows such as form filling, data extraction from dashboards, or testing software interfaces where visual state changes drive the next action.
How does tool calling work in this model?+
Tool calling enables the model to trigger predefined functions — like clicking a button, typing text, or retrieving data from an API — based on its analysis of visual input, allowing it to act within digital environments.
How does this model differ from other OpenAI vision models?+
Unlike models focused only on describing images, Computer Use Preview is optimized for reasoning about interface states and planning sequences of actions to achieve a goal in interactive software.
What should users know about its current limitations?+
As a preview release, it may struggle with highly dynamic or low-contrast interfaces, and its performance depends on the clarity and stability of the visual input provided.
Compare with others
Org chart: how to move your company onto AI
A practical map: which company roles and processes AI agents can take over, where to start, and in what order to roll it out.