Claude API Flask Example

Build a Claude chatbot backend with Flask in Python. Includes streaming with SSE using stream_with_context, conversation history endpoint, and CORS setup.

Flask is the most widely deployed Python web framework. This guide shows how to build a Claude chatbot backend with Flask — from minimal synchronous calls to full server-sent event (SSE) streaming.

Installation

pip install flask anthropic flask-cors

Minimal synchronous endpoint

from flask import Flask, request, jsonify
import anthropic

app = Flask(__name__)
client = anthropic.Anthropic()  # reads ANTHROPIC_API_KEY from env

@app.route("/chat", methods=["POST"])
def chat():
    user_message = request.json.get("message", "")
    response = client.messages.create(
        model="claude-sonnet-4-6",
        max_tokens=1024,
        messages=[{"role": "user", "content": user_message}]
    )
    return jsonify({"reply": response.content[0].text})

if __name__ == "__main__":
    app.run(debug=True)

Test it:

curl -X POST http://localhost:5000/chat \
  -H "Content-Type: application/json" \
  -d '{"message": "What is prompt caching?"}'

Streaming with SSE (server-sent events)

from flask import Flask, request, Response, stream_with_context
import anthropic, json

app = Flask(__name__)
client = anthropic.Anthropic()

@app.route("/chat/stream", methods=["POST"])
def chat_stream():
    user_message = request.json.get("message", "")

    def generate():
        with client.messages.stream(
            model="claude-sonnet-4-6",
            max_tokens=1024,
            messages=[{"role": "user", "content": user_message}]
        ) as stream:
            for text in stream.text_stream:
                yield f"data: {json.dumps({'text': text})}\n\n"
        yield "data: [DONE]\n\n"

    return Response(
        stream_with_context(generate()),
        mimetype="text/event-stream",
        headers={
            "Cache-Control": "no-cache",
            "X-Accel-Buffering": "no"  # disable nginx buffering
        }
    )

Consume the SSE stream in JavaScript

const source = new EventSource('/chat/stream');  // GET version
// Or for POST with fetch:
const resp = await fetch('/chat/stream', {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({message: 'Hello Claude'})
});
const reader = resp.body.getReader();
const decoder = new TextDecoder();
while (true) {
  const {done, value} = await reader.read();
  if (done) break;
  const lines = decoder.decode(value).split('\n');
  for (const line of lines) {
    if (line.startsWith('data: ') && line !== 'data: [DONE]') {
      const {text} = JSON.parse(line.slice(6));
      document.getElementById('output').textContent += text;
    }
  }
}

Conversation history endpoint

from flask import Flask, request, jsonify, session
import anthropic

app = Flask(__name__)
app.secret_key = "change-me-in-production"
client = anthropic.Anthropic()

@app.route("/chat/history", methods=["POST"])
def chat_with_history():
    if "history" not in session:
        session["history"] = []

    user_message = request.json.get("message", "")
    session["history"].append({"role": "user", "content": user_message})

    response = client.messages.create(
        model="claude-sonnet-4-6",
        max_tokens=1024,
        messages=session["history"]
    )

    assistant_reply = response.content[0].text
    session["history"].append({"role": "assistant", "content": assistant_reply})
    session.modified = True  # required: Flask won't detect nested mutations

    return jsonify({"reply": assistant_reply, "turns": len(session["history"]) // 2})

@app.route("/chat/reset", methods=["POST"])
def reset():
    session.pop("history", None)
    return jsonify({"ok": True})

CORS setup for a separate frontend

from flask_cors import CORS

app = Flask(__name__)
CORS(app, origins=["http://localhost:3000", "https://yourapp.com"])

Flask vs FastAPI for Claude streaming

Estimate token costs for your expected traffic at the Claude API Cost Calculator. For async patterns see the FastAPI streaming guide and async Python patterns.

Frequently asked questions

Free tools

More examples