Show HN: Dereference.dev – Prompt-First IDE for Parallel Claude Code Sessions

3 months ago 3

Orchestrate multiple AI sessions. Branch conversations atomically. Built for developers who ship fast.

Download other versions

For the lovers of

See it in action

Experience the power of AI-driven development with our intuitive interface.

Parallel AI Sessions

Run multiple AI conversations simultaneously. Switch between Claude, GPT-4, and Gemini instantly.

Branch & Merge

Create branches from any conversation point. Explore different solutions and merge the best results.

Multi-Session Orchestration

Run multiple AI sessions in parallel. Switch contexts instantly.

Manage Claude, GPT-4, and Gemini sessions simultaneously. Each session maintains its own context, allowing you to compare approaches, validate solutions, and leverage the strengths of different models for different tasks.

Atomic Branching

Branch and merge conversations like code. Never lose context.

Create branches at any point in your conversation history. Explore alternative solutions without losing your original context. Merge successful branches back into your main conversation flow, just like Git.

Lightning Fast

Native performance with instant response times.

Built with Rust and optimized for speed. Experience near-instant responses, smooth animations, and efficient memory usage. No electron overhead - pure native performance on every platform.

Privacy First

Your code stays local. No telemetry. No tracking.

All processing happens on your machine. We don't collect usage data, track your code, or send analytics. Your API keys are stored securely in your local keychain. Complete privacy by design.

AI Tool Integration

Works with Claude Code, Gemini CLI, and more.

Seamlessly integrates with popular AI coding assistants. Use your existing tools and workflows while gaining the power of multi-session orchestration and branching. Compatible with any LLM that provides an API.

Smart Context

Intelligent context management across all sessions.

Automatically manages context windows for optimal performance. Smart truncation ensures relevant information stays in focus while maintaining conversation coherence across branches and sessions.