Cloudflare Workers AI

Verified ✅ Truly Free

Run open models on Cloudflare's global network. Workers AI includes 10,000 free neurons per day on every account, no credit card required, shared across the free catalog: GPT-OSS 120B and 20B, Qwen3.8 27B, Gemma 4 26B, Llama 3.3 70B and 4 Scout, Mistral Small 3.1, Nemotron 3 Super, plus Whisper and FLUX image generation. The newest GLM-5.x, Kimi K2.6/K2.7 and DeepSeek V4 models now require a paid billing method. Requests need your account ID in the URL.

Free Tier Edge Computing Global Network No Credit Card
Get API Key Suggest Edit
6

Overview

Provider Type

API

API Endpoint

https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/run/

Free Tier Highlights

Varies by model

Why Choose Cloudflare Workers AI?

Every Cloudflare account includes 10,000 free "neurons" per day for Workers AI, on the free plan too and with no credit card. That budget covers a broad catalog of free models, from GPT-OSS 120B, Qwen3.8 27B and Llama 3.3 70B to Whisper speech-to-text and FLUX image generation, all on Cloudflare's global network.

Best for: a reliable free catalog for apps and side projects, including speech and images.

Watch out: you need your account ID in the URL, the newest GLM-5.x, Kimi and DeepSeek V4 models require a paid billing method, and 10,000 neurons a day goes faster on large models.

Quick Start Guide

1

Sign up at https://dash.cloudflare.com/

2

Go to AI > Workers AI

3

Get your Account ID from the dashboard

4

Create an API Token with Workers AI permissions

5

Use the REST API or Workers bindings

Available Models

22 free models
Model Name ID Context Capabilities
FLUX.2 Dev (image) Free
@cf/black-forest-labs/flux-2-dev
- -
Whisper Large v3 Turbo Free
@cf/openai/whisper-large-v3-turbo
- -
EuroLLM 9B Free
@cf/utter-project/eurollm-9b-it
- -
Apertus v1.5 8B Free
@cf/swiss-ai/apertus-v1.5-8b
- -
Granite 4.0 H Micro Free
@cf/ibm-granite/granite-4.0-h-micro
- -
Gemma SEA-LION v4 27B Free
@cf/aisingapore/gemma-sea-lion-v4-27b-it
- -
DeepSeek R1 Distill Qwen 32B Free
@cf/deepseek-ai/deepseek-r1-distill-qwen-32b
- -
Qwen2.5 Coder 32B Free
@cf/qwen/qwen2.5-coder-32b-instruct
- -
QwQ 32B Free
@cf/qwen/qwq-32b
- -
Llama 3.2 11B Vision Free
@cf/meta/llama-3.2-11b-vision-instruct
- -
Llama 3.1 8B Instruct (fp8) Free
@cf/meta/llama-3.1-8b-instruct-fp8
- -
Llama 3.3 70B Instruct (fp8 fast) Free
@cf/meta/llama-3.3-70b-instruct-fp8-fast
24 000 Context
-
Nemotron 3 Super 120B A12B Free
@cf/nvidia/nemotron-3-120b-a12b
256 000 Context
-
Qwen3.8 27B Free
@cf/qwen/qwen3.8-27b
262 000 Context
-
GLM 4.7 Flash Free
@cf/zai-org/glm-4.7-flash
Long context
-
Mistral Small 3.1 24B Free
@cf/mistralai/mistral-small-3.1-24b-instruct
128 000 Context
-
Qwen3 30B A3B Free
@cf/qwen/qwen3-30b-a3b-fp8
32 000 Context
-
Gemma 4 26B A4B Free
@cf/google/gemma-4-26b-a4b-it
Long context
-
Llama 4 Scout 17B Free
@cf/meta/llama-4-scout-17b-16e-instruct
131 000 Context
-
GPT OSS 20B Free
@cf/openai/gpt-oss-20b
128 000 Context
-
GPT OSS 120B Free
@cf/openai/gpt-oss-120b
128 000 Context
-
Llama 3.2 3B Instruct Free
@cf/meta/llama-3.2-3b-instruct
128 000
-

Integration Examples

Ready-to-use code snippets for your applications.

main.py
import requests

ACCOUNT_ID = "YOUR_ACCOUNT_ID"
API_TOKEN = "YOUR_API_TOKEN"

response = requests.post(
    f"https://api.cloudflare.com/client/v4/accounts/{ACCOUNT_ID}/ai/run/@cf/meta/llama-3.1-8b-instruct",
    headers={"Authorization": f"Bearer {API_TOKEN}"},
    json={
        "messages": [
            {"role": "user", "content": "What is Workers AI?"}
        ]
    }
)

print(response.json()["result"]["response"])

Free Tier Pricing & Limits

Rate Limit

Requests per minute

Varies by model

Daily Quota

Requests per day

10000 neurons/day

Token Limit

Tokens per minute

Included in neuron budget

Monthly Quota

Per month limit

~300000 neurons/month

Use Cases

Serverless AI applications

Edge-first AI products

Low-latency global chatbots

Image classification and generation

Content moderation at the edge

Limitations & Considerations

10,000 neurons/day cap (varies per model)

Larger models consume more neurons per request

No fine-tuning support

Some models have limited context windows

Beta models may change

Community Hub

Live

Join the discussion, share tips, and rate Cloudflare Workers AI.

Quick Reactions

Add Discussion

Comments are moderated. Be helpful and respectful.

Recent Activity

0 comments

Ready to Get Started?

Join thousands of developers using Cloudflare Workers AI

Start Building Now