Perplexity API Simple

A simple, clean drop-in replacement for the official Perplexity API. Perfect for using with TaskMaster AI and Claude Code MCP server.

Features

• ✅ 100% Perplexity-focused - No multi-provider complexity
• ✅ OpenAI-compatible API - Works with any OpenAI SDK
• ✅ Cookie authentication - Use your Perplexity account for free searches
• ✅ Web dashboard - Manage API keys and monitor usage
• ✅ Hot-reload cookies - Update cookies without restarting
• ✅ Token tracking - Monitor your usage

Quick Start

1. Installation

cd /home/anombyte/projects/perplexity-api-simple

# Install dependencies
pip install -r requirements.txt

# Create .env file
cat > .env <<EOF
PERPLEXITY_COOKIE=your-cookie-here-optional
EOF

2. Start the Server

# Simple start
./scripts/start_server.sh

# Or manually
export $(cat .env | xargs) && python src/perplexity_api_server.py

The server will start at: http://localhost:8765

3. Generate API Key

1. Open http://localhost:8765 in your browser
2. Click "Generate New API Key"
3. Copy your API key (starts with pplx_)

Usage

For TaskMaster AI

Add to your TaskMaster configuration:

# In your .env or shell profile
export PERPLEXITY_API_KEY="pplx_your-generated-key-here"
export PERPLEXITY_API_BASE_URL="http://localhost:8765"

Then use TaskMaster normally with the --research flag:

tm parse-prd --research
tm expand-all --research

For Claude Code (MCP Server)

1. Generate API Key from the web dashboard at http://localhost:8765

2. Update your MCP settings at ~/.claude/mcp_config.json:

{
  "mcpServers": {
    "perplexity-free": {
      "command": "node",
      "args": [
        "/path/to/perplexity-api-free/mcp-server/build/index.js"
      ],
      "env": {
        "PERPLEXITY_API_KEY": "pplx_your-generated-key-here",
        "PERPLEXITY_API_BASE_URL": "http://localhost:8765"
      }
    }
  }
}

3. Restart Claude Code

4. Test the connection:

   • Ask Claude: "Use Perplexity to search for the latest Python best practices"
   • Claude should use the MCP tool automatically

For Any OpenAI-Compatible Client

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:8765/v1",
    api_key="pplx_your-generated-key-here"
)

response = client.chat.completions.create(
    model="sonar-pro",
    messages=[
        {"role": "user", "content": "What are the latest developments in AI?"}
    ]
)

print(response.choices[0].message.content)

Available Models

• sonar - Default Perplexity model (fast)
• sonar-pro - Pro model (more accurate)
• sonar-reasoning - Reasoning model (deep thinking)
• sonar-deep-research - Deep research mode

API Endpoints

• GET / - Web dashboard
• POST /chat/completions - Perplexity API endpoint
• POST /v1/chat/completions - OpenAI-compatible endpoint
• GET /models - List available models
• GET /health - Health check

Cookie Setup (Optional)

To use your Perplexity account for authenticated searches:

Getting Your Cookie

1. Login to Perplexity at https://www.perplexity.ai
2. Open DevTools (F12)
3. Go to Application/Storage → Cookies
4. Copy the entire cookie string

Option 1: Via Web Dashboard

1. Open http://localhost:8765
2. Click "Cookie Settings"
3. Paste your cookie
4. Click "Save" (hot-reloads without restart!)

Option 2: Via .env File

# Edit .env file
PERPLEXITY_COOKIE=your-full-cookie-string-here

Then restart the server.

Troubleshooting

Server Won't Start

# Check if port 8765 is already in use
lsof -i :8765

# Kill existing process
kill -9 $(lsof -t -i :8765)

# Try again
./scripts/start_server.sh

API Key Not Working

1. Open http://localhost:8765
2. Check if your key is active in the dashboard
3. Regenerate a new key if needed

Cookie Not Working

1. Make sure you're logged into Perplexity.ai
2. Copy the ENTIRE cookie string (not just one field)
3. Try refreshing your Perplexity session
4. Use the web dashboard to update the cookie (hot-reload)

WSL-Specific Issues

If running in WSL and can't access from Windows:

# Make sure you're using 0.0.0.0 (not 127.0.0.1)
# Server already binds to 0.0.0.0 by default

# Access from Windows using WSL IP
ip addr show eth0 | grep inet
# Use: http://<wsl-ip>:8765

Development

Project Structure

perplexity-api-simple/
├── src/
│   ├── perplexity_api_server.py  # Main server
│   └── perplexity_fixed.py       # Perplexity client
├── web/
│   └── index.html                # Dashboard
├── scripts/
│   └── start_server.sh           # Startup script
├── requirements.txt
├── .env.example
└── README.md

Running in Development

# With auto-reload
FLASK_DEBUG=1 python src/perplexity_api_server.py

# With custom port
PORT=9000 python src/perplexity_api_server.py

Docker Support

# Build image
docker build -t perplexity-api-simple .

# Run container
docker run -p 8765:8765 \
  -e PERPLEXITY_COOKIE="your-cookie" \
  -v $(pwd)/.api_keys.json:/app/.api_keys.json \
  perplexity-api-simple

Security

• API keys are stored locally in .api_keys.json
• Cookies are stored in .env (never committed to git)
• The server runs locally and doesn't expose your keys to the internet
• Use HTTPS in production deployments

Contributing

This is a simplified, Perplexity-focused version. For multi-provider support, see the main perplexity-api-free project.

License

MIT

Credits

Based on the perplexity-api-free project, simplified for ease of use with TaskMaster and Claude Code.