CodeGPT

CodeGPT is an AI coding assistant for developers using Visual Studio Code or JetBrains IDEs. It reviews a codebase, proposes an editable plan, and helps implement code changes using selected AI models.

One of 61 tools in Coding

CodeGPT screenshot

Who it's for

  • Software developers using Visual Studio Code who want an AI assistant for code generation, refactoring, or debugging.
  • Developers working in JetBrains IDEs, including IntelliJ or PyCharm, who want an AI coding extension.
  • Engineering teams that want pooled credits, centralized billing, and usage statistics for their coding assistant.

How it fits your workflow

Install the CodeGPT extension in Visual Studio Code or a JetBrains IDE, then choose whether to use your own API keys or CodeGPT credits and select a model.

CodeGPT reads the codebase, proposes a plan that you can edit, and suggests code changes. You approve steps before changes are made; terminal commands also wait for approval before anything mutates.

Pricing

Free: A free plan is available with usage and feature limits.

Free plan

$0

  • · Free and paid plans are documented. Feature and usage limits depend on the selected plan.

Paid plans

Price not published

  • · See the provider pricing page for current rates, billing periods, and usage limits.

Prices checked on Sep 7, 2026 from the vendor's site. They can change; confirm before you buy.

Key features

Bring Your Own Key (BYOK)
Use your own API keys and run any supported model; CodeGPT emphasizes that you own the keys and control when cloud models are used.
Model Agnostic / Full Model Flexibility
Supports choosing or swapping models (mentions Claude, Gemini, OpenAI, Anthropic, etc.) to match reasoning depth, context length, or speed requirements.
IDE Integrations
Extensions for Visual Studio Code and JetBrains IDEs (IntelliJ, PyCharm, etc.) to bring AI coding workflows directly into developer environments.
AI Agents & Planning Mode
Agent and Planning mode for multi-step reasoning and agent-driven workflows that plan and execute coding tasks.
MCP (Integration Bridge)
Links to APIs, documentation, and databases to allow the assistant and agents to access external systems and resources.
Context Tracking & Rules
Tracks context live with rules the assistant follows for consistent, standards-based code and transparent context usage without hidden limits.

Works with

  • Visual Studio Code
  • JetBrains IDEs
  • Model providers (OpenAI, Anthropic, Claude, Gemini, etc.)
  • MCP - APIs, Docs, Databases

Limitations to know

  • The page says code remains under the user's control unless they explicitly choose to use cloud AI models; using those models can involve sending code to cloud AI services.

Alternatives to consider

  • GIDE, Your AI Coding Command Center

    Choose it if you want a local-first coding IDE and CLI that runs models on your machine, works offline, and offers an OpenAI-compatible local API.

Compare CodeGPT side by side

Getting started

  1. Download the extension for Visual Studio Code or a JetBrains IDE.
  2. Choose your model access method: bring your own API keys or use CodeGPT credits.
  3. Select a model and ask CodeGPT to work with your codebase.
  4. Review and approve the proposed plan, edits, or terminal actions.