Lynkr

An LLM Gateway which optimises your token usage.

87.6% fewer tokens on JSON tool results. 53% fewer tokens on tool-heavy requests. 171ms semantic cache hits. Zero code changes.

87.6% JSON Compression

53% Tool Token Reduction

171ms Semantic Cache Hits

13+ LLM Providers

0 Code Changes Required

Numbers from a live benchmark against LiteLLM on identical workloads. See full report →

---

🚀 New: Wrap Mode for AI Coding Tools

Use Lynkr's routing with your AI coding assistant — maximize your subscription value:

npm install -g lynkr


# Claude Code Pro/Max
lynkr wrap claude

Wrapping gives you:

• ✅ Tier routing (send simple tasks to free Ollama, complex to your subscription/API)
• ✅ TOON/RTK compression (87% token reduction on tool outputs)
• ✅ Semantic caching (171ms cache hits)
• ✅ 3-5x more usage from the same subscription limits
• ✅ Works with OAuth (Claude, Copilot, Cursor) or API keys (Aider, Codex)

Full wrap guide →

---

Quick Start (2 Minutes)

1. Install Lynkr

npm install -g lynkr

2. Configure Lynkr

The fastest path is the interactive wizard:

lynkr init

It asks four questions — usage mode (Claude Pro/Max via wrap, or direct API keys), tier picks for SIMPLE/MEDIUM/COMPLEX/REASONING across the 12 supported providers, credentials for what you chose, and a few routing-intelligence knobs — then writes a fully-populated .env with sensible production defaults for everything else (caching, compression, policy budgets, MCP sandbox, agents, rate limiting).

Useful flags:

lynkr init --force                # overwrite an existing .env

See docs/init.md for the full wizard reference.

If you'd rather configure by hand, the manual options below still work — copy .env.example to .env and edit it directly:

Option A: Free & Local (Ollama) - Recommended for Testing

# Install Ollama first: https://ollama.com
ollama pull qwen2.5-coder:latest

Then start Lynkr:

lynkr start

3. Connect Your Tool

Cursor IDE

• Settings → Models → Override Base URL
• Set to: http://localhost:8081/v1
• API Key: any-value

Codex CLI

Edit ~/.codex/config.toml:

model_provider = "lynkr"

[model_providers.lynkr]
base_url = "http://localhost:8081/v1"
wire_api = "responses"

✅ Done! Your AI tool now uses your chosen provider.

---

Common Startup Errors

Error: unable to determine transport target for "pino-pretty"

Problem: You're running an older version (< 9.3.0).

Solution: Update to the latest version:

npm install -g lynkr@latest

If you must use an older version, set NODE_ENV=production before starting.

Warning: Missing tier configuration: TIER_SIMPLE, TIER_MEDIUM...

This is just a warning - you can ignore it. Tier routing is optional.

To remove the warning, add to .env:

TIER_SIMPLE=ollama:qwen2.5-coder:latest
TIER_MEDIUM=ollama:qwen2.5-coder:latest
TIER_COMPLEX=ollama:qwen2.5-coder:latest
TIER_REASONING=ollama:qwen2.5-coder:latest

Warning: FALLBACK_PROVIDER='databricks' is enabled but missing credentials

Solution: Add to .env:

FALLBACK_ENABLED=false

Error: connect ECONNREFUSED ::1:11434 (Ollama)

Problem: Ollama is not running.

Solution:

ollama serve

Keep this terminal open, and start Lynkr in a new terminal.

Error: Connection refused or 404 Not Found

Problem: Lynkr is not running or wrong port.

Solution: Check Lynkr is running on the correct port:

curl http://localhost:8081/

Should return: {"service":"Lynkr","version":"9.x.x","status":"running"}

---

Why Lynkr?

AI coding tools lock you into one provider and send every token raw. Lynkr breaks both locks.

Claude Code / Cursor / Codex / Cline / Continue
                    ↓
                  Lynkr
          ┌─────────────────────┐
          │  Strip unused tools  │  ← 53% fewer tokens on tool calls
          │  Compress JSON blobs │  ← 87.6% on large tool results
          │  Semantic cache      │  ← 171ms hits, 0 tokens billed
          │  Route by complexity │  ← cheap model for simple, cloud for hard
          └─────────────────────┘
                    ↓
    Ollama | Bedrock | Azure | Moonshot | OpenRouter | OpenAI

What you get:

• ✅ 53% fewer tokens on tool-heavy requests (Claude Code, Cursor sessions)
• ✅ 87.6% compression on large JSON tool results (grep, file reads, test output)
• ✅ Semantic cache serves repeated queries in 171ms with 0 tokens billed
• ✅ Automatic tier routing — simple questions go to cheap models, complex ones escalate
• ✅ Route through your company's infrastructure (Databricks, Azure, Bedrock)
• ✅ Zero code changes — just change one environment variable

---

Supported Providers

Provider	Type	Example Models	Cost
Ollama	Local	qwen2.5-coder, deepseek-coder, llama3	Free
llama.cpp	Local	Any GGUF model	Free
LM Studio	Local	Local models with GUI	Free
OpenRouter	Cloud	GPT-4o, Claude 3.5, Llama 3, Gemini	$
AWS Bedrock	Cloud	Claude, Llama, Mistral, Titan	$$
Databricks	Cloud	Claude Sonnet 4.5, Opus 4.6	$$$
Azure OpenAI	Cloud	GPT-4o, o1, o3	$$$
Azure Anthropic	Cloud	Claude Sonnet, Opus	$$$
OpenAI	Cloud	GPT-4o, o3-mini	$$$
DeepSeek	Cloud	DeepSeek R1, Reasoner	$

4 local providers for 100% offline, free usage. 10+ cloud providers for scale.

---

Advanced: Tier Routing (Save Even More)

Route different request types to different models automatically:

# .env file
MODEL_PROVIDER=ollama
FALLBACK_ENABLED=false

# Use small/fast models for simple tasks
TIER_SIMPLE=ollama:qwen2.5:3b

# Use medium models for normal coding
TIER_MEDIUM=ollama:qwen2.5:7b

# Use powerful models for complex architecture
TIER_COMPLEX=ollama:deepseek-r1:14b
TIER_REASONING=ollama:deepseek-r1:14b

# Optional: Limits (remove for unlimited) for long conversations
POLICY_MAX_STEPS=50
POLICY_MAX_TOOL_CALLS=100

Lynkr analyzes each request and routes it to the appropriate tier. Simple questions use fast models. Complex refactoring uses powerful models.

Result: 70-90% of requests use cheaper/faster models. Only hard problems hit expensive models.

Tier configuration is strictly authoritative — bandit exploration is constrained to the models you've listed in TIER_*, and multi-turn conversations score with a recency-weighted sliding window so context isn't lost on short follow-ups. See docs/intent-window-routing.md.

---

Common Issues & Fixes

Issue	Solution
"Service temporarily overloaded"	Ollama model too large for RAM. Use smaller model or increase --max-old-space-size
"Route not found: HEAD /"	Ignore - harmless health check from Claude Code
"Hallucinated tool calls"	Normal - Lynkr automatically filters invalid tools
"Safe Command DSL blocked"	Add POLICY_SAFE_COMMANDS_ENABLED=false to .env
"spawn graphify ENOENT"	Optional feature. Set CODE_GRAPH_ENABLED=false in .env (see Advanced Features section for installation)
Slow first request (20+ sec)	Ollama loading model into memory. Add OLLAMA_KEEP_ALIVE=30m in Ollama config
No response after N turns	Remove POLICY_MAX_STEPS and POLICY_MAX_TOOL_CALLS from .env (unlimited by default in v9.3.0+)

---

Advanced Features

Token Optimization (60-80% savings)

# Enable all optimizations
PROMPT_CACHE_ENABLED=true
SEMANTIC_CACHE_ENABLED=true
TOOL_INJECTION_ENABLED=false
CODE_MODE_ENABLED=true

Always-on (no config): smart tool selection (server mode), RTK tool-result compression (test/git/grep/lint/build/JSON output), MCP tool dedup (drops built-in WebSearch/WebFetch when an Exa/Tavily MCP tool is present), and request bypass (Claude CLI Warmup / title-extraction calls are answered locally, never hitting a provider).

Optional terse-output mode to cut output tokens:

CAVEMAN_ENABLED=true        # off by default — nudges the model to be concise
CAVEMAN_LEVEL=lite          # lite | full | ultra

Cost tracking & model pricing

Per-request cost is computed from a model-pricing registry (LiteLLM → models.dev, cached 24h) and recorded in telemetry. Models the registry doesn't know record cost_usd=null (logged once) rather than a fabricated price. Pin prices for unknown models:

# Per-1M-token USD prices, JSON keyed by model name
MODEL_PRICE_OVERRIDES={"my-model":{"input":0.5,"output":1.5}}

Memory System (Titans-inspired)

MEMORY_ENABLED=true
MEMORY_TTL=3600000  # 1 hour

Load Shedding & Resilience

LOAD_SHEDDING_ENABLED=true
LOAD_SHEDDING_HEAP_THRESHOLD=0.85

Admin Hot-Reload (no restart needed)

curl -X POST http://localhost:8081/v1/admin/reload

Code Intelligence (Optional - Graphify)

Graphify provides AST-based code analysis for smarter routing decisions.

Installation (Rust required):

# Install Rust if not already installed
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
source $HOME/.cargo/env

# Build and install graphify
git clone https://github.com/safishamsi/graphify
cd graphify
cargo build --release
sudo cp target/release/graphify /usr/local/bin/

# Verify installation
graphify --version

Enable in .env:

CODE_GRAPH_ENABLED=true
CODE_GRAPH_WORKSPACE=/path/to/your/project  # Optional, defaults to cwd

Features:

• AST-based complexity scoring
• Structural code analysis (19 languages supported)
• Enhanced routing decisions based on code structure

Note: Graphify is completely optional. If not installed, Lynkr falls back to simpler complexity analysis.

---

Installation Methods

NPM (recommended)

npm install -g lynkr

One-line installer

curl -fsSL https://raw.githubusercontent.com/Fast-Editor/Lynkr/main/install.sh | bash

Homebrew (macOS / Linux)

brew tap fast-editor/lynkr
brew install lynkr
lynkr --version

Upgrade later with brew update && brew upgrade lynkr. The formula tracks the latest lynkr npm release automatically.

Docker

git clone https://github.com/Fast-Editor/Lynkr.git
cd Lynkr
docker-compose up -d

From source

git clone https://github.com/Fast-Editor/Lynkr.git
cd Lynkr
npm install
cp .env.example .env
npm start

---

Documentation

Guide	Description
Installation	All installation methods
Provider Setup	Configuration for all 12+ providers
Claude Code	Claude Code CLI integration
Cursor IDE	Cursor setup + troubleshooting
Codex CLI	Codex configuration
Tier Routing	Smart model routing by complexity
Token Optimization	60-80% cost reduction
Troubleshooting	Common issues and solutions
API Reference	REST API endpoints
Production	Enterprise deployment

---

Benchmark Results

Head-to-head against LiteLLM on the same backends (Ollama minimax-m2.5, Moonshot, Azure OpenAI), 9 scenarios across 4 feature categories. Apples-to-apples comparison is Lynkr vs LiteLLM billed tokens on the same scenario. Run with node benchmark-tier-routing.js.

Run: June 5, 2026 · Lynkr v9.3.2 · LiteLLM v1.87.1 · macOS, Apple Silicon.

Token reduction (vs LiteLLM, same model & prompt)

Mechanism	Lynkr	LiteLLM	Result
Smart tool selection (14 tools)	959 tokens · $0.0044	2,085 tokens · $0.0091	53% fewer tokens, 52% cheaper
TOON compression (60-item grep JSON)	427 tokens · $0.009	3,458 tokens · $0.018	87.6% fewer tokens, 50% cheaper

Lynkr strips irrelevant tool schemas (smart tool selection) and binary-compresses large JSON tool results (TOON) — both in-process, no added latency.

Semantic cache

	Tokens billed	Response time
First call (cold)	2,857	1,891ms
Second call — paraphrased, cache hit	0 (served from cache)	171ms (11× faster)

Near-identical prompts return cached responses in 171ms. Zero model tokens billed on a cache hit.

Tier routing

Request	Lynkr routes to	LiteLLM routes to
"What does git stash do?"	minimax-m2.5 (local, free)	Ollama (local)
JWT vs cookies security analysis	moonshot (cloud — correct)	Ollama (local — wrong call)

Lynkr scores each request on 15 dimensions (token count, code complexity, reasoning markers, risk signals, agentic patterns) and escalates automatically. LiteLLM's cost-based-routing sends everything to the cheapest model regardless of complexity.

Cost projection (100,000 requests/month, same backend)

	Monthly cost	vs LiteLLM
LiteLLM	~$818	baseline
Lynkr	~$409	~50% cheaper

Based on a tool-heavy agentic session (TOON scenario). On equal footing — same provider, same model — Lynkr is cheaper due to token optimization.

→ Full benchmark report with methodology

---

Cost Comparison

Scenario	Direct Anthropic	Lynkr + Ollama	Lynkr + OpenRouter
Daily coding (8h)	$10-30/day	$0 (free)	$2-8/day
Monthly (heavy use)	$300-900	$0	$60-240

With tier routing + token optimization: additional 50-87% savings on cloud providers depending on workload.

---

Why Lynkr vs Alternatives

Feature	Lynkr	LiteLLM	OpenRouter	PortKey
Setup	npm install -g lynkr	Python + Docker + Postgres	Account signup	Docker stack
Claude Code native	✅ Drop-in	⚠️ Requires config	❌	⚠️ Partial
Cursor native	✅ Drop-in	⚠️ Partial	❌	⚠️ Partial
Local models	Ollama, llama.cpp, LM Studio	Ollama only	❌	❌
Automatic tier routing	✅ 15-dimension scorer	⚠️ Cost-only	❌	❌ Manual metadata
TOON JSON compression	✅ up to 87.6%	❌	❌	❌
Smart tool selection	✅ up to 60% token reduction	❌	❌	❌
Semantic cache	✅ 171ms hits, 0 tokens	❌	❌	✅ Prompt cache only
Long-term memory	✅ SQLite, per-session	❌	❌	❌
MCP integration	✅ + Code Mode (96% reduction)	❌	❌	❌
Self-hosted	✅ Node.js only	✅ Python stack	❌ SaaS	✅ Docker
Dependencies	Node.js 20+	Python, Prisma, PostgreSQL	None	Docker, Python

Lynkr's edge: Purpose-built for AI coding tools. Compresses tokens before they reach the model — not just after. Zero-config for Claude Code, Cursor, and Codex. Installs in one command.

---

Community

• GitHub Discussions — Ask questions
• Report Issues — Bug reports
• NPM Package — Official releases
• DeepWiki — AI-powered docs

---

License

Apache 2.0 — See LICENSE.

---

Built by Vishal Veera Reddy for developers who want control over their AI tools.