Self-hosted proxy server that enables Claude Code, Cursor, and AI coding tools to work with any LLM provider — with 60-80% cost reduction.
Enterprise-grade features for teams that demand reliability, performance, and cost control.
Seamlessly switch between Ollama, AWS Bedrock, OpenRouter, Databricks, Azure, and more — without changing a single line of code.
Smart tool selection, prompt caching, and memory deduplication work together to reduce your LLM costs by 60-80%.
Titans-inspired memory system persists context across sessions, enabling smarter and more contextual responses.
Complete tool calling support with Model Context Protocol for advanced agent workflows and custom integrations.
Built-in circuit breakers, load shedding, metrics, health checks, and audit logging for production deployments.
Connect to models on any machine — run inference on dedicated GPU servers and share resources across your entire team.
Intelligent request routing across model tiers — automatically route simple tasks to faster models and complex ones to more capable models.
From free local models to enterprise cloud solutions — use what works best for your use case.
Drop-in replacement for Anthropic API — no code changes required in any of your tools.
Three simple steps to unlock any LLM provider with your favorite coding tools.
# Using npm
npm install -g lynkr
# Or using Homebrew
brew tap vishalveerareddy123/lynkr && brew install lynkrexport MODEL_PROVIDER=ollama
export OLLAMA_MODEL=qwen2.5-coder:latestexport MODEL_PROVIDER=bedrock
export AWS_BEDROCK_API_KEY=your-api-key
export AWS_BEDROCK_MODEL_ID=anthropic.claude-3-5-sonnet-20241022-v2:0# Start the proxy server
npm start # → Running at http://localhost:8081
# Connect Claude Code CLI
export ANTHROPIC_BASE_URL=http://localhost:8081
export ANTHROPIC_API_KEY=dummy
claudeLynkr sits between your tools and any LLM provider, handling format conversion, caching, and optimization transparently.
Claude Code / Cursor / Codex / ClawdBot
│
▼
┌─────────────────────┐
│ Lynkr Proxy │
│ localhost:8081 │
│ │
│ • Format Conversion│
│ • Prompt Caching │
│ • Token Optimization
│ • Tool Calling │
│ • Memory System │
│ • Metrics & Logs │
└──────────┬──────────┘
│
┌───────────────┼───────────────┐
▼ ▼ ▼
┌─────────┐ ┌─────────┐ ┌─────────┐
│ Local │ │ Cloud │ │ Remote │
├─────────┤ ├─────────┤ ├─────────┤
│ Ollama │ │ Bedrock │ │GPU Server│
│llama.cpp│ │OpenRouter │ (any IP)│
│LM Studio│ │Databricks │ │
│MLX Server │Azure/OpenAI │ │
└─────────┘ └─────────┘ └─────────┘
Join developers saving 60-80% on LLM costs while using their favorite coding tools with any provider.