perplexity-webui-scraper 1.0.0

An advanced, high-performance Python client, MCP server, and REST API for reverse-engineering Perplexity AI's WebUI.

Perplexity WebUI Scraper

Python scraper to extract AI responses from Perplexity's web interface.

---

📚 Full Documentation & Advanced Guide: https://henrique-coder.github.io/perplexity-webui-scraper

---

What is this?

This library lets you interact with Perplexity AI programmatically using the same web endpoints as the browser — no official API key required. It supports conversations, file uploads, streaming, an MCP server for AI agents, and a drop-in OpenAI-compatible REST API.

• Requirements: A Perplexity Pro or Max account and your browser session token.
• Key Features: 15 models (GPT-5.4, Claude Opus, Gemini, Deep Research…), file attachments (images, PDFs, …), streaming, MCP Server for AI agents, OpenAI-compatible REST API, multi-turn conversation thread continuation.

Installation

# Core library only
uv add perplexity-webui-scraper

# Interactive session token generator (adds rich)
uv add "perplexity-webui-scraper[cli]"

# MCP Server for AI agents (adds fastmcp)
uv add "perplexity-webui-scraper[mcp]"

# OpenAI-compatible API server (adds fastapi + uvicorn + typer)
uv add "perplexity-webui-scraper[api]"

# Everything at once
uv add "perplexity-webui-scraper[cli,mcp,api]"

Quick Start

1. Get your session token

# Interactive CLI wizard — walks you through email auth
uv run get-perplexity-session-token

Or retrieve __Secure-next-auth.session-token manually from your browser cookies on perplexity.ai.

2. Basic usage

from perplexity_webui_scraper import Perplexity

client = Perplexity(session_token="YOUR_TOKEN")
conversation = client.create_conversation()

conversation.ask("What is quantum computing?")
print(conversation.answer)

# Follow-ups preserve context automatically
conversation.ask("Explain it simpler")
print(conversation.answer)

3. Streaming

for chunk in conversation.ask("Explain AI", stream=True):
    if chunk.last_chunk:
        print(chunk.last_chunk, end="", flush=True)

4. Choose a model

from perplexity_webui_scraper import ConversationConfig

conversation = client.create_conversation(ConversationConfig(model="perplexity/best"))
conversation.ask("Solve this step by step: ...")
print(conversation.answer)

5. List all available models

from perplexity_webui_scraper import MODELS

for model in MODELS.list_all():
    print(f"{model.id:40} {model.name}")

Available CLIs

Command	Extra	Description
get-perplexity-session-token	cli	Interactive email auth wizard to generate a session token
perplexity-webui-scraper-mcp	mcp	Start the MCP server (used via MCP config, not directly)
perplexity-webui-scraper-api	api	Start the OpenAI-compatible REST API server

OpenAI-Compatible API

Run a local server that accepts OpenAI-formatted requests and forwards them to Perplexity. Works as a drop-in replacement for any OpenAI client — authentication is done per-request via Authorization: Bearer, exactly like the real API.

# Start the server (no token needed at startup)
perplexity-webui-scraper-api

# Custom host and port
perplexity-webui-scraper-api --host 0.0.0.0 --port 8080

# Development mode with auto-reload
perplexity-webui-scraper-api --reload

Running via Container (Podman / Docker)

You can seamlessly run the REST API using the provided Containerfile via Podman or Docker. This is the recommended way to securely isolate the server. The project utilizes a modern Python 3.14 Alpine container powered by uv for lightning-fast builds.

# 1. Build the lightweight image
podman build -t perplexity-api .

# 2. Run the server (exposing port 8000)
podman run --rm -it -p 8000:8000 perplexity-api

> You can safely replace podman with docker in the commands above as the Containerfile is fully OCI-compatible.

CLI options

Option	Short	Default	Description
--host	-H	127.0.0.1	Bind address
--port	-p	8000	Port to listen on
--reload		False	Enable auto-reload (dev)
--log-level		info	Uvicorn log level

Authentication

Pass your Perplexity session token as the API key in every request — exactly like the OpenAI API:

# curl
curl http://localhost:8000/v1/chat/completions \
  -H "Authorization: Bearer YOUR_SESSION_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"model": "perplexity/best", "messages": [{"role": "user", "content": "Hello!"}]}'

# Streaming
curl -N http://localhost:8000/v1/chat/completions \
  -H "Authorization: Bearer YOUR_SESSION_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"model": "perplexity/best", "messages": [{"role": "user", "content": "Hello!"}], "stream": true}'

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:8000/v1",
    api_key="YOUR_SESSION_TOKEN",  # sent as Authorization: Bearer automatically
)

response = client.chat.completions.create(
    model="perplexity/best",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

API endpoints

Method	Path	Description
GET	/v1/models	List all available models
POST	/v1/chat/completions	Chat completion (streaming + non-streaming)
GET	/docs	Interactive Swagger UI
GET	/redoc	ReDoc documentation

Fields not supported by Perplexity (e.g. temperature, top_p) are accepted for client compatibility but silently ignored.

MCP Server

Expose every Perplexity model as a separate tool for AI agents (Claude Desktop, Antigravity, etc.):

{
  "mcpServers": {
    "perplexity-webui-scraper": {
      "command": "uvx",
      "args": [
        "--from",
        "perplexity-webui-scraper[mcp]@latest",
        "perplexity-webui-scraper-mcp"
      ],
      "env": { "PERPLEXITY_SESSION_TOKEN": "your_token_here" }
    }
  }
}

See the full MCP documentation for all tools and configuration details.

Disclaimer

This is an unofficial library. It uses internal APIs that may change without notice. Use at your own risk. By using this library, you agree to Perplexity AI's Terms of Service.