Skip to content
  • Models
  • Rankings
  • Ori
Sign Up
Sign Up
OpenRouterOpenRouter
© 2026 OpenRouter, Inc

Product

  • Chat
  • Rankings
  • Benchmarks
  • Apps
  • Discover
  • Models
  • Collections
  • Providers
  • Tools
  • Pricing
  • Business
  • Enterprise
  • Labs

Company

  • About
  • Blog
  • Careers
    Hiring
  • Privacy
  • Terms of Service
  • Trust Center
  • Support
  • Works With OR
  • Data
  • Brand

Developer

  • Documentation
  • API Reference
  • Developer Platform
  • Status
  • AI Site Map

Connect

  • Discord
  • GitHub
  • LinkedIn
  • X
  • YouTube
All server tools

Apply patch

openrouter:apply_patch

Docs

Lets coding agents edit files with structured patches. The model emits a V4A diff, OpenRouter validates it and returns a structured apply_patch call, and your application applies the change to its own files.

Docs
How it worksProviders and pricingUsing this toolRelated tools

How it works

The building block for coding agents that edit files.

  1. 1

    Your model writes a V4A diff that creates, updates, or deletes a file.

  2. 2

    OpenRouter validates the diff, and returns invalid patches to the model to fix.

  3. 3

    Your app applies the patch to its own files and reports the result on the next turn.

Providers and pricing

  • Favicon for https://openrouter.ai

    OpenRouter validator

    Free

    OpenRouter validates the V4A diff and returns a structured apply_patch call.

  • Favicon for https://openai.com

    OpenAI

    Native
    Provider pricing

    Built-in apply_patch on GPT-5 and later.

    Via OpenAI, Azure

Standard model token costs apply to every request.

Using this tool

OpenRouter validates the diff and returns it as an apply_patch_call output item. Your application applies the patch to its own files and sends the result back as apply_patch_call_output on the next turn.

Available through the API. Patches are applied by your application, so there is no file system to edit in the chatroom.

Supported APIs: Responses.

shell
curl https://openrouter.ai/api/v1/responses \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "openai/gpt-5.2",
  "tools": [
    {
      "type": "openrouter:apply_patch"
    }
  ],
  "input": "Rename the function greet to sayHello in src/index.ts."
}'

Setup, parameters, and examples in the docs.

Read the integration guide

Related tools

Code & execution

  • Shell

    A sandboxed hosted shell for any model

    Code & execution

  • Bash

    Anthropic-style bash tool with optional sandboxed execution

    Code & execution