"Lintasan adalah jalur tempat koneksi, kecerdasan, dan kemungkinan bergerak."

Di dunia modern, AI bukan hanya tentang model. AI adalah tentang bagaimana sistem saling terhubung, bagaimana request menemukan jalur terbaik, dan bagaimana manusia berinteraksi dengan teknologi secara efisien.

Lintasan hadir sebagai jalur cerdas yang menghubungkan manusia, AI, dan sistem dalam satu aliran yang terintegrasi. Satu endpoint untuk semua provider AI — tidak perlu lagi berganti-ganti SDK atau mengelola banyak API key di berbagai tempat.

"Lintasan is the path where connections, intelligence, and possibilities move."

In the modern world, AI isn't just about models. It's about how systems connect, how requests find their best path, and how humans interact with technology efficiently.

Lintasan exists as an intelligent pathway connecting humans, AI, and systems in one integrated flow. One endpoint for all AI providers — no more switching SDKs or managing multiple API keys across different places.

Lintasan adalah LLM proxy gateway dengan 40+ fitur optimasi. Satu endpoint OpenAI-compatible untuk semua provider AI — routing cerdas, embedded cache, dual-mode CommandCode, plugin system, dan penghematan token otomatis.

Masalah yang diselesaikan:

• 🔀 Multi-provider complexity — ganti provider = ganti SDK, ganti API key, ganti format
• 💸 No cost visibility — tidak tahu berapa token terpakai per model/provider
• 🔄 No failover — satu provider down, semua request gagal
• 🔐 Key management chaos — API key tersebar di mana-mana
• 📊 No observability — tidak bisa lihat usage, latency, error rate

Solusi Lintasan:

• Satu endpoint → semua provider (OpenAI, Anthropic, DeepSeek, Gemini, Groq, dll)
• Satu API key → autentikasi terpusat
• Smart routing → request otomatis ke provider terbaik
• Circuit breaker → provider gagal auto-disable
• Dashboard → monitoring real-time 17 halaman
• Plugin system → extensible tanpa ubah core

Lintasan is an LLM proxy gateway with 40+ optimization features. One OpenAI-compatible endpoint for all AI providers — smart routing, embedded cache, dual-mode CommandCode, plugin system, and automatic token savings.

Problems solved:

• 🔀 Multi-provider complexity — switching providers means switching SDKs, API keys, and formats
• 💸 No cost visibility — can't see token usage per model/provider
• 🔄 No failover — one provider down, all requests fail
• 🔐 Key management chaos — API keys scattered everywhere
• 📊 No observability — can't see usage, latency, error rates

Lintasan's solution:

• One endpoint → all providers (OpenAI, Anthropic, DeepSeek, Gemini, Groq, etc.)
• One API key → centralized authentication
• Smart routing → requests automatically go to the best provider
• Circuit breaker → failing providers auto-disable
• Dashboard → 17-page real-time monitoring
• Plugin system → extensible without core changes

Opsi 1: Binary Pre-built (Recommended)

Satu binary, dashboard sudah termasuk di dalamnya.

curl -L -o lintasan https://github.com/sanhaji182/lintasan-go/releases/latest/download/lintasan-linux-amd64
chmod +x lintasan
./lintasan start

Opsi 2: Build dari Source

Butuh Go 1.22+ dan Node 20+ (untuk build dashboard). make build meng-compile frontend SvelteKit jadi SPA statis, meng-embed-nya ke binary Go, lalu compile binary tunggal.

git clone https://github.com/sanhaji182/lintasan-go.git
cd lintasan-go
make build       # frontend → embed → ./lintasan
./lintasan start

Tanpa Node? go build -o lintasan ./cmd/lintasan tetap jalan, tapi menghasilkan server API-only (tanpa UI dashboard).

Opsi 3: Docker (single container)

git clone https://github.com/sanhaji182/lintasan-go.git
cd lintasan-go
LINTASAN_MASTER_KEY=$(openssl rand -hex 32) docker compose up --build
# UI + API → http://localhost:20180

CLI Commands

./lintasan start      # Start server (UI + API, default :20180)
./lintasan setup      # Initialize database
./lintasan mitm start # Start MITM HTTPS bridge (optional)
./lintasan version    # Show version
./lintasan help       # All commands

# Custom port
PORT=8080 ./lintasan start

Option 1: Pre-built Binary (Recommended)

One binary, dashboard included.

curl -L -o lintasan https://github.com/sanhaji182/lintasan-go/releases/latest/download/lintasan-linux-amd64
chmod +x lintasan
./lintasan start

Option 2: Build from Source

Requires Go 1.22+ and Node 20+ (to build the dashboard). make build compiles the SvelteKit frontend into a static SPA, embeds it into the Go binary, and compiles a single executable.

git clone https://github.com/sanhaji182/lintasan-go.git
cd lintasan-go
make build       # frontend → embed → ./lintasan
./lintasan start

No Node? go build -o lintasan ./cmd/lintasan still works but produces an API-only server (no dashboard UI).

Option 3: Docker (single container)

git clone https://github.com/sanhaji182/lintasan-go.git
cd lintasan-go
LINTASAN_MASTER_KEY=$(openssl rand -hex 32) docker compose up --build
# UI + API → http://localhost:20180

CLI Commands

./lintasan start      # Start server (UI + API, default :20180)
./lintasan setup      # Initialize database
./lintasan mitm start # Start MITM HTTPS bridge (optional)
./lintasan version    # Show version
./lintasan help       # All commands

# Custom port
PORT=8080 ./lintasan start

Tidak perlu .env file — set env vars atau gunakan default. Database auto-create saat pertama run.

No .env file needed — just set env vars or use defaults. Database auto-creates on first run.

Lintasan dilengkapi dashboard interaktif berbasis SvelteKit 5 + Tailwind v4 untuk monitoring dan konfigurasi real-time.

Akses: http://localhost:20180/dashboard (via nginx reverse-proxy ke SvelteKit di port 5173)

Lintasan comes with an interactive dashboard built with SvelteKit 5 + Tailwind v4 for real-time monitoring and configuration.

Access: http://localhost:20180/dashboard (via nginx reverse-proxy to SvelteKit at port 5173)

Major Providers (4)

OpenAI · Anthropic · DeepSeek · Google Gemini

Top-Tier (6)

xAI (Grok) · Mistral AI · Azure OpenAI · Azure AI Foundry · Google Vertex AI · AWS Bedrock

AI Coding (4)

Codestral API · GitHub Copilot API · Pydantic AI Agents · Meta Llama API

Aggregators (8)

OpenRouter · Replicate · HuggingFace Inference · Vercel AI Gateway · AIML API · Poe by Quora · CometAPI · NanoGPT

High-Speed Inference (10)

Groq · Together AI · Fireworks AI · Cerebras · NVIDIA NIM · Cloudflare Workers AI · Hyperbolic · Lambda AI · FriendliAI · Anyscale Endpoints

GPU Cloud (12)

Baseten · OctoAI · Lepton AI · Featherless AI · Crusoe Cloud · nscale AI · PublicAI · Galadriel · Chutes · GMI Cloud · Heroku AI · Novita AI

CommandCode (2)

CommandCode (API Key) · CommandCode (Alpha)

Chinese Providers (10)

GLM / Zhipu AI · Kimi / Moonshot · MiniMax · Qwen / Alibaba · SiliconFlow · Xiaomi MiMo · Volcano Engine (ByteDance) · Z.AI · DeepSeek · Baidu Qianfan

Indonesia Providers (2)

Sumopod · Apertis AI (Stima API)

Enterprise & Cloud (8)

Snowflake Cortex AI · Oracle Cloud (OCI) · SAP AI Core · IBM watsonx · Gradient AI · NLP Cloud · Petals · Clarifai

Specialized (20)

Perplexity · Cohere · DeepInfra · SambaNova · Nebius AI · Aleph Alpha · AI21 Labs · Reka AI · Voyage AI · Deepgram · Black Forest Labs · Stability AI · Runway ML · Recraft AI · fal.ai · Helicone · Lemonade AI · Bytez · Sarvam AI · MorphDB

Self-Hosted (3)

Ollama · vLLM · LM Studio

Other (24)

AWS SageMaker · GigaChat · Predibase · OpenPipe · Scale AI · Titan ML · OctoML · Monster API · GooseAI · Forefront · Custom

# Build binary
cd lintasan-go
go build -ldflags="-s -w" -o lintasan ./cmd/lintasan

# Build frontend
cd frontend && npm install && npm run build && cd ..

# Copy to server
scp lintasan user@server:/opt/lintasan/
scp -r frontend/build user@server:/opt/lintasan/frontend-build/

# Systemd service
sudo tee /etc/systemd/system/lintasan.service << 'EOF'
[Unit]
Description=Lintasan LLM Proxy Gateway
After=network.target

[Service]
Type=simple
User=lintasan
WorkingDirectory=/opt/lintasan
ExecStart=/opt/lintasan/lintasan start
Restart=on-failure
RestartSec=5

[Install]
WantedBy=multi-user.target
EOF

sudo systemctl enable --now lintasan

Nginx Reverse Proxy

server {
    listen 443 ssl;
    server_name your-domain.com;

    # API → Go backend
    location /api/ { proxy_pass http://127.0.0.1:20180; }
    location /v1/  { proxy_pass http://127.0.0.1:20180; }

    # Dashboard → SvelteKit
    location / { proxy_pass http://127.0.0.1:5173; }
}

Lintasan Go menggunakan database schema yang berbeda dari Node.js v1 — tidak backward-compatible.

1. Export settings dari Node v1 dashboard
2. Install Lintasan Go (lihat Instalasi)
3. Re-create connections di Go v2 dashboard atau via API
4. Sync models — auto-discovered saat pertama connect
5. Verifikasi — test lewat Playground

⚠️ Repo Node.js v1 sudah di-archive di sanhaji182/lintasan

Lintasan Go uses a different database schema from Node.js v1 — not backward-compatible.

1. Export settings from Node v1 dashboard
2. Install Lintasan Go (see Installation)
3. Re-create connections in Go v2 dashboard or via API
4. Sync models — auto-discovered on first connection
5. Verify — test via Playground

⚠️ Node.js v1 repo is archived at sanhaji182/lintasan

Kontribusi sangat diterima! Buka issue untuk bug report atau feature request.

1. Fork repo
2. Buat branch (git checkout -b feature/amazing-feature)
3. Commit changes (git commit -m 'feat: amazing feature')
4. Push ke branch (git push origin feature/amazing-feature)
5. Buka Pull Request

Pastikan tests tetap passing:

go test ./...     # Harus 0 failures

Contributions welcome! Open an issue for bug reports or feature requests.

1. Fork the repo
2. Create a branch (git checkout -b feature/amazing-feature)
3. Commit changes (git commit -m 'feat: amazing feature')
4. Push to branch (git push origin feature/amazing-feature)
5. Open a Pull Request

Make sure tests still pass:

go test ./...     # Must be 0 failures

Project ini dibangun dengan AI sebagai development partner. Arsitektur, keputusan teknis, dan quality control tetap di tangan manusia — AI mengeksekusi.

Lintasan Go adalah rewrite total dari Node.js v1 — mempertahankan semua fitur dengan footprint 35x lebih ringan. Dibuat untuk komunitas AI Indonesia yang percaya bahwa masa depan development adalah orkestrasi, bukan sekadar coding manual.

This project was built with AI as a development partner. Architecture, technical decisions, and quality control remain in human hands — AI executes.

Lintasan Go is a complete rewrite of Node.js v1 — retaining all features with 35x lighter footprint. Made for the AI community that believes the future of development is orchestration, not just manual coding.