computer-use-preview — Deprecated
- Deprecated
- —
- Shutdown
- 2026-07-23
- Status
- deprecated
- Replacement
- 5.4-mini
Quick fix — copy & paste
Choose your language. The "before" block matches the deprecated call; the "after" block is the drop-in replacement.
response = client.responses.create(
model="computer-use-preview",
tools=[{"type": "computer_use_preview", "display_width": 1024, "display_height": 768}],
input=[{"role": "user", "content": "Open a browser"}],
) response = client.responses.create(
model="gpt-5", # use a regular model + the GA computer use tool
tools=[{"type": "computer_use", "display_width": 1024, "display_height": 768}],
input=[{"role": "user", "content": "Open a browser"}],
) const response = await openai.responses.create({
model: "computer-use-preview",
tools: [{ type: "computer_use_preview", display_width: 1024, display_height: 768 }],
input: [{ role: "user", content: "Open a browser" }],
}); const response = await openai.responses.create({
model: "gpt-5",
tools: [{ type: "computer_use", display_width: 1024, display_height: 768 }],
input: [{ role: "user", content: "Open a browser" }],
}); "model": "computer-use-preview", "type": "computer_use_preview" "model": "gpt-5", "type": "computer_use" Error messages
Seeing one of these? You're in the right place.
-
model_not_found: computer-use-preview -
model_not_found: computer-use-preview-2025-03-11
Replacement options
-
5.4-miniCompare token cost →
Also known as
These ids point to the same deprecated model.
computer-use-preview-2025-03-11 Other OpenAI deprecations
What this means for your code
computer-use-preview is an agentic API surface (Assistants, computer-use, or similar) with its own request/response shape. Agentic APIs are not just model ids — they are entire request flows. Migration means rewriting middleware: thread management, tool definitions, polling vs streaming, and state persistence.
OpenAI has scheduled computer-use-preview for shutdown on 2026-07-23. That gives you 76 days to migrate. Until then the model still works, but every API call after that date will return a model_not_found error.
Find every call in your codebase
Before you change anything, locate every place the deprecated model id is referenced. Search source files, environment files, feature flags, and config repos. Use these commands from your project root:
Python projects
grep -rn '"computer-use-preview"' --include="*.py" . JavaScript / TypeScript projects
grep -rn '"computer-use-preview"' --include="*.{js,ts,tsx,jsx}" . Anywhere (configs, scripts, infra)
grep -rn "computer-use-preview" . Run the same searches for each alias listed above (computer-use-preview-2025-03-11). Different aliases for the same model often coexist in older code paths.
Migration checklist
Steps in order. Skip any that don't apply, but read the whole list — for agent models, the non-obvious steps are usually the ones that break in production.
- 1. Map the deprecated API's lifecycle events to the replacement API's
- 2. Move thread/conversation state from provider-managed to your own store, if needed
- 3. Re-implement tool definitions in the new API's schema
- 4. Update polling, streaming, and webhook handlers
- 5. Run shadow traffic against the new API before cutover; budget 1-3 days of engineering
Will this migration cost more?
Switching from computer-use-preview to 5.4-mini could change your costs significantly. Calculate the exact difference for your prompts.
Open the cost calculator →