CodeGPT: AI Coding Agents & Chat installs statistics

CodeGPT: Chat & AI Agents

<!-- ═══ THE FRONTIER SLOT ═════════════════════════════════════════════ The first thing every visitor sees. Rotate it with each frontier model drop: new hero PNG in code-gpt-docs/assets/, new headline, keep the "try it free" mechanics honest (the model must be free via the daily Economy quota — no `pro` flag in api-codegpt-deno credits-menu.json — or spendable on welcome credits). 2026-09-10: DeepSeek V4.1 Flash — an ECONOMY model (no `pro` flag), so the honest CTA is "free every day": it rides the 10-a-day quota and is the menu default. Restored 2026-10-05 after OpenRouter retired the stealth Space Bunny Alpha (hero-space-bunny-alpha.png, 09-24 to 10-05). Previous: hero-gpt-6-astra-death-star.png (GPT-6 Astra, 2026-09-04, premium — "try it with our free credits"); hero-claude-fable-5-1-king.png (Claude Fable 5.1, 2026-09-01). --> <img alt="DeepSeek V4.1 Flash — into the unknown. A frontier-class model, free every day in CodeGPT" src="https://raw.githubusercontent.com/JudiniLabs/code-gpt-docs/main/assets/hero-deepseek-v4-1-flash-spirit-v2.png" width="100%">

Into the unknown. DeepSeek V4.1 Flash is in CodeGPT today, and it is free. A 552B model with a new encoder-decoder architecture that reaches 98% of GPT-6 Astra's score at 1.4% of the price, with a 1M-token context and full agent mode. It is now the default model in CodeGPT: free every day on the daily quota, no card needed. Just open the chat and start.

<!-- ═══ END FRONTIER SLOT ═════════════════════════════════════════════ -->

And every day is free: all accounts get 10 free AI interactions daily — no trial clock. Use Claude, GPT, Gemini, or local models, with agents that create, edit, and read files directly in your project. No lock-in. Your choice of provider, your way of coding.

---

Free, Every Day

Sign in and start coding — no payment details, ever, on the free tier.

• 10 chat & agent interactions every day on fast Economy models — DeepSeek V4.1 Flash, GLM 5.3 Flash, Gemini 3.8 Flash, GPT 5.6 Luna, GLM 5.2, and more. Resets daily.
• $1.50 in welcome credits to spend on any premium model — including frontier models like GPT-6 Astra and Claude Fable 5.1.
• Bring your own API keys or run local models — unlimited, free forever.

---

Agent Mode

AI agents that go beyond chat. They create files, apply edits, read your codebase, and execute multi-step tasks — all from a single prompt.

<img alt="Agent Mode" src="https://raw.githubusercontent.com/JudiniLabs/code-gpt-docs/main/assets/explain-tool-execution.gif" width="100%">

---

Smart Diff

Review every AI-suggested change with a real-time diff preview before applying. Accept, reject, or edit — you stay in control.

<img alt="Smart Diff" src="https://raw.githubusercontent.com/JudiniLabs/code-gpt-docs/main/assets/editFile-diff-implementation.gif" width="100%">

---

Chat

Ask questions, debug issues, get explanations, and generate code — with full context from your project files.

<img alt="Chat" src="https://raw.githubusercontent.com/JudiniLabs/code-gpt-docs/main/assets/chat_landing.gif" width="100%">

---

Autocomplete

Fast, context-aware code completions powered by your choice of AI model.

<img alt="Autocomplete" src="https://raw.githubusercontent.com/JudiniLabs/code-gpt-docs/main/assets/autocomplete_landing.gif" width="100%">

---

Top Models. No API Keys Needed.

Use the best models without any setup — new frontier models are added the day they launch.

Claude Fable 5.1 · Claude Opus 5 · Claude Sonnet 5 · GPT-5.6 · Gemini 3.8 Flash · DeepSeek V4.1 Flash · Kimi K3 · and more

<img alt="Add Credits" src="https://raw.githubusercontent.com/JudiniLabs/code-gpt-docs/main/assets/add-credits-flux.gif" width="100%">

---

Simple Plans

Free — $0	10 AI interactions every day · $1.50 welcome credits to spend on any premium model · unlimited with your own API keys or local models
Professional — $10/mo	Unlimited Economy model usage · unlimited autocomplete · access to premium models — Claude, GPT, Gemini — with a monthly allowance for premium tasks · top up any time
Teams — $30/seat/mo	Everything in Professional · pooled credits across your team · centralized billing · admin dashboard

See full pricing →

---

Bring Your Own Key

Connect directly to 20+ AI providers. Your keys, direct API calls, no intermediaries.

OpenAI · Anthropic · Google · Mistral · Groq · DeepSeek · Azure · Bedrock · Nvidia · Cohere · Fireworks · Perplexity · Cerebras · Grok · and more

---

Local Models

Run models locally with Ollama, LM Studio, or Docker. Complete privacy — zero data leaves your machine.

<img alt="Local Models" src="https://raw.githubusercontent.com/JudiniLabs/code-gpt-docs/main/assets/local-models-modal.png" width="100%">

---

Chat Commands

Command	Description
/Fix	Detect and fix errors in your code
/Explain	Get a clear explanation of selected code
/Refactor	Refactor for readability and maintainability
/Document	Generate documentation
/Unit Testing	Generate unit tests
#file-name	Add files as context for the AI

---

Get Started…