Puter Pool — Free AI API with No Limits 🚀

Get a free AI API for Claude Opus 5, GPT-5.6 Sol, Gemini 3.7 Flash, DeepSeek V4, Qwen, Llama, Mistral, Grok and 400+ models by pooling free Puter accounts.
Unlimited credits — no card, no rate limits, no token costs.

🌐 Live Landing • 🛠️ Local Depot Dashboard • 🔌 Local API • 📦 Repository

Features • Quick Start • How It Works • API • Dashboard • Docker • Testing

---

How it works

Puter gives every free account a daily AI credit allowance. Puter Pool combines the allowances of accounts you own into a single OpenAI-compatible (/v1/chat/completions) and Anthropic-compatible (/v1/messages) endpoint, so coding agents (opencode, Claude Code, etc.) keep running without ever hitting a cap.

Notes

Puter may ask for phone verification when creating accounts. Only pool accounts you own, and respect Puter's terms of service.

✨ Features

Feature	Description
Free AI API	OpenAI at /v1/chat/completions + Anthropic at /v1/messages, /anthropic/v1/messages (all SSE streaming)
400+ Models	Live from Puter — Claude Opus 5, GPT-5.6 Sol, Gemini 3.7 Flash, DeepSeek V4, Mistral, Qwen, Grok, Llama… grouped by lab with LobeHub icons
Unlimited Credits	Automatic failover across accounts — never hit a cap
Local by Design	Depot runs on localhost only; tokens never leave your machine. Landing is static on Cloudflare Workers
Industrial UI	Depot dashboard with P logo (#0C0F12 ink + #FF3B1F signal + #FDB813 bolt), rail conveyor, health rail, queue meter, request tape, live traffic chart
Smart Scheduler	Round-robin or least-used strategy, retries up to MAX_RETRIES=3, parks exhausted/error accounts via HealthChecker
Local SQLite	data/pool.db via node:sqlite — auto-created, zero config
Docker Ready	docker compose up -d persists ./data + ./logs

---

🎬 Demo

Watch the dashboard in action: account rack, live chat, health rail, queue meter, and request tape.

---

🚀 Quick Start

Prerequisites

• Node.js 20+
• npm or pnpm
• (Optional) Docker & Docker Compose

1. Clone & Install

git clone https://github.com/Parithosh-Varma/puter-pool.git
cd puter-pool
cp .env.example .env   # SQLite at data/pool.db by default — no config needed
npm install && npm run build && npm start

2. Start Dashboard (separate terminal)

cd dashboard && npm install && npm run dev
# → Tool: http://localhost:5173 (no auth, localhost-only)
# → API:  http://localhost:3000  (landing stays on Cloudflare)

3. Add Puter Accounts

1. Create free accounts at puter.com
2. Get your token from browser DevTools → Application → Local Storage → puter.com → auth_token
3. Paste tokens in the Depot UI at http://localhost:5173

---

⚙️ How It Works

flowchart LR
    A[Your App] --> B[Puter Pool API :3000]
    B --> C{Scheduler}
    C --> D[Account 1]
    C --> E[Account 2]
    C --> F[Account N]
    D --> G[Puter API]
    E --> G
    F --> G
    G --> H[AI Response]
    H --> A

Loading

1. Create multiple Puter accounts (free at puter.com) — each gives daily AI credits
2. Paste tokens in Depot at http://localhost:5173 — health + credit tracked per account
3. Point any SDK at one endpoint — OpenAI POST /v1/chat/completions or Anthropic POST /v1/messages
4. Scheduler shunts — round-robin or least-used, retries up to MAX_RETRIES=3, parks exhausted/error bays via HealthChecker
5. Uninterrupted — pool handles rotation transparently, tape logs every request

---

🔌 API Reference

Proxy Endpoints (Same Pool, Same Failover)

Method	Endpoint	Format	Description
POST	/v1/chat/completions	OpenAI	{model, messages, stream, max_tokens, temperature}
POST	/v1/messages	Anthropic	{model, max_tokens, messages, system, temperature, stream}
POST	/anthropic/v1/messages	Anthropic (alt)	SDK baseURL: "/anthropic"
POST	/anthropic/messages	Anthropic (alt)
GET	/v1/models	OpenAI	Live model list from Puter
GET	/healthz	—	Liveness probe
GET	/readyz	—	Readiness (503 if 0 active accounts)

App Endpoints

Method	Endpoint	Description
POST	/api/ai/chat	Auto-routed {model, prompt, stream}
GET	/api/models	Puter models (RSS + details)
GET	/api/accounts	List with health + credit (token redacted)
POST	/api/accounts	Add account {name, token, dailyCreditLimit}
DELETE	/api/accounts/:id	Remove account
PATCH	/api/accounts/:id	Update {status, token, name}
GET	/api/stats	Pool + scheduler stats
GET	/api/dashboard	Full UI data payload
GET	/api/history?limit=100	Request history

---

Quick Examples

OpenAI curl

curl http://localhost:3000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-5","messages":[{"role":"user","content":"Hello"}]}'

Anthropic curl

curl http://localhost:3000/v1/messages \
  -H "Content-Type: application/json" \
  -H "x-api-key: ignored" -H "anthropic-version: 2023-06-01" \
  -d '{"model":"claude-opus-5","max_tokens":1024,"messages":[{"role":"user","content":"Hello"}]}'

OpenAI SDK (Python)

from openai import OpenAI
client = OpenAI(base_url="http://localhost:3000/v1", api_key="ignored")
print(client.chat.completions.create(
    model="gpt-5.6-sol",
    messages=[{"role":"user","content":"Hello"}]
).choices[0].message.content)

Anthropic SDK (Python)

import anthropic
c = anthropic.Anthropic(base_url="http://localhost:3000", api_key="ignored")
print(c.messages.create(
    model="claude-opus-5",
    max_tokens=1024,
    messages=[{"role":"user","content":"Hello"}]
).content[0].text)

---

⚙️ Configuration

Full options in .env.example. Key variables:

Variable	Default	Description
PORT	3000	API server port
SQLITE_PATH	data/pool.db	Local SQLite file (auto-created)
SCHEDULER_STRATEGY	round-robin	round-robin | least-used
MAX_RETRIES	3	Accounts to try before returning 5xx
HEALTH_CHECK_INTERVAL_MS	60000	Re-verify account health interval
REQUEST_TIMEOUT_MS	30000	Per-request timeout
LOG_LEVEL	info	Logging verbosity

---

🛠️ Dashboard

http://localhost:5173 — TOOL · LOCALHOST ONLY (no login required)

• 💬 Chat with any model (streaming)
• 📦 Account Rack — add/enable/disable/delete cartridges
• 📊 Metrics — health / credit / latency / errorRate / consecutiveFailures
• 🚂 Queue Rail (128 slots) + live tape + traffic chart
• ⚙️ Strategy Switch — round-robin ↔ least-used

---

🗄️ Database

Local SQLite at SQLITE_PATH (data/pool.db) via node:sqlite — auto-created on first run.

• Tables: puter_accounts + ai_responses
• Mounted in Docker at ./data
• Reference schema: supabase-schema.sql

---

🐳 Docker

docker compose up -d   # uses data/pool.db, no env needed

Persists:

• ./data → SQLite database
• ./logs → Application logs

---

☁️ Landing Page (Separate Deploy)

Static Vite app at /landing (gitignored), deployed to Cloudflare Workers:

cd landing && npm install && npm run build
source ~/bin/load-env.zsh  # CLOUDFLARE_API_TOKEN + CLOUDFLARE_ACCOUNT_ID
npx wrangler deploy        # → https://puterpool.parithosh.workers.dev

Features: Hero with curl examples, 400+ models grouped by lab (live RSS), how-it-works animation, API docs, comparison table, local steps, FAQ — LobeHub CDN icons + custom depot SVGs.

---

🧪 Testing

npm test              # 54 tests (Vitest)
npm run test:coverage # With coverage report
npm run typecheck     # TypeScript strict check
npm run build         # Verify compilation

---

📁 Project Structure

puter-pool/
├── src/                 # Core TypeScript modules
│   ├── accounts/        # Account management
│   ├── scheduler/       # Round-robin / least-used
│   ├── health/          # HealthChecker
│   ├── proxy/           # OpenAI + Anthropic proxies
│   ├── db/              # SQLite layer
│   └── index.ts         # Entry point
├── dashboard/           # React + Vite depot UI
├── landing/             # Cloudflare Workers landing (gitignored)
├── tests/               # Vitest suites
├── data/                # SQLite (mounted in Docker)
├── logs/                # Application logs
├── assets/              # Demo video, screenshots (gitignored)
└── logo.png             # Brand assets

Star History

---

🤝 Contributing

1. Fork the repo
2. Create a feature branch: git checkout -b feat/amazing-feature
3. Run checks: npm run typecheck && npm test && npm run build
4. Commit: git commit -m 'feat: add amazing feature'
5. Push and open a PR

---

📄 License

Apache-2.0 — see LICENSE for details.

---

🙏 Acknowledgements

• Puter — for the free AI credits
• LobeHub — for provider icons
• Vite + React — for the dashboard
• Cloudflare Workers — for the landing deployment