Cloudflare gives you 10,000 free Neurons per day on Workers AI, with a daily reset at 00:00 UTC. This Cloudflare Workers AI free tier is genuinely free: no credit card required, and there’s no time limit.
It’s a solid deal for devs shipping a prototype, small startup teams watching burn, and researchers who need inference without waiting on a GPU box. You can run LLMs, embeddings, image generation, and speech models on Cloudflare’s edge network.
Below: eligibility, the exact signup paths (dashboard, REST API, or Wrangler), the limits that matter, and a few practical ways to stretch your 10,000 Neurons further.
Program at a Glance
| Provider | Cloudflare |
| Credit Amount | 10,000 Neurons/day (resets daily) |
| Duration | No time limit; resets 00:00 UTC |
| Eligibility | Anyone with a Cloudflare account on Workers Free/Paid |
| Credit Card Required? | No, not for the free plan |
| Difficulty | Easy; instant signup and start in dashboard |
| Best For | Edge inference, prototypes, full-stack AI demos |
| Official Page | Cloudflare Program Page |
What You Actually Get
Workers AI includes a daily allowance of 10,000 Neurons to run AI models on Cloudflare’s global edge network. Neurons are Cloudflare’s way of measuring GPU compute across different model types, including LLMs, image generation, embeddings, speech-to-text, and text-to-speech. Cloudflare also supports OpenAI-compatible endpoints for chat completions, text completions, and embeddings, which means you can often use the standard OpenAI SDK by changing the base URL. On top of that, Workers AI hosts 100+ open-source models, and popular models are designed to avoid the “cold start” pain you might be used to.
In practical terms, one day’s free allowance can translate to roughly 1,300 LLM responses, about 12,500 embeddings, around 2,000 small image generations (example: FLUX.1 Schnell at 512×512), or about 8,300 image classifications. Those are estimates, not guarantees, because neuron cost varies a lot by model size. Still, it’s enough to build a real demo, run a small internal tool, or power a lightweight production feature if you’re careful.
Who Qualifies (and Who Doesn’t)
Eligibility is refreshingly simple: if you can create a Cloudflare account, you can use Workers AI. The 10,000 Neurons/day apply to Workers Free plan users and Workers Paid plan users.
- You need a Cloudflare account created via the Cloudflare dashboard signup flow.
- Email verification is required before you can reliably use the dashboard features.
- To use the REST API path, you must create a Workers AI API token with Workers AI Read and Workers AI Edit permissions.
- If you use Wrangler locally (including
wrangler dev), it still consumes your account’s neuron quota.
There aren’t the usual “new customers only” traps here. The main “don’t qualify” scenario is more practical than legal: once you hit 10,000 Neurons on the Free plan, Cloudflare hard-blocks additional usage until the daily reset.
How to Sign Up
Signup is quick, and you can start either in the dashboard, via REST API, or with Wrangler.
- Go to dash.cloudflare.com/sign-up and create a free Cloudflare account (email + password, no credit card).
- Verify your email address.
- In the dashboard, navigate to Workers & Pages.
- Pick how you want to use Workers AI: Dashboard app template, REST API, or the Wrangler CLI.
- For REST API usage, open Workers AI, click Use REST API, then create a Workers AI API Token and copy your Account ID.
After that, you’re effectively live. Credits aren’t something you “apply” for; the daily neuron allowance is just there, and usage shows up in Workers AI analytics in the Cloudflare dashboard.
What the Credits Cover
Your 10,000 Neurons/day can be spent across model types, not just chat. Workers AI covers text generation, embeddings, image generation, speech-to-text, text-to-speech, and a grab bag of other tasks like translation, summarization, reranking, and safety classification.
| Service / Feature | What It Does | Included? |
|---|---|---|
| LLMs (text generation) | Chat and text completions on open-source models. | ✓ |
| Embeddings endpoint | Vector embeddings for search and RAG pipelines. | ✓ |
| Image generation | Generate images (example models include FLUX and SDXL variants). | ✓ |
| OpenAI-compatible API | Use OpenAI SDK patterns by swapping base URL. | ✓ |
Notable exclusions are mostly about limits, not access. On the Free plan, after 10,000 Neurons you get a hard block (an error is returned), and separate per-model request-per-minute caps still apply even if you have neurons left.
Limitations to Know About
Every free program has catches. Cloudflare’s are straightforward, but you should know them before you build something that gets real traffic.
- The Free plan has a hard stop at 10,000 Neurons/day, and requests return an error after you hit it.
- Per-model rate limits apply separately from neurons, typically around 1,500–3,000 requests per minute depending on the model.
- Model availability varies by Cloudflare data center, so some less-used models may route to fewer locations with slightly higher latency.
- If you use Workers AI locally through
wrangler dev, it still burns your daily neuron quota.
When the daily free neurons run out, what happens depends on your plan. Workers Free users are blocked until the next reset at 00:00 UTC. Workers Paid users still get the same 10,000 free neurons, but can continue past the cap with overflow billing at about $0.011 per 1,000 neurons.
Have Unused Cloudflare Credits?
Free tiers are great, but bigger Cloudflare programs (startup and enterprise deals) can leave teams sitting on credits they won’t realistically use before the clock runs out. It happens a lot: roadmap changes, infra moves, or the “AI feature” gets de-prioritized. If you’ve got surplus Cloudflare credits, AI Credit Mart lets you sell unused allocations instead of letting them expire.
Need More Cloudflare Credits?
Once you outgrow the 10,000 Neurons/day, you don’t necessarily need to pay full retail to keep experimenting. AI Credit Mart lists discounted Cloudflare credits from organizations that can’t use them all. Typical discounts land around 30–70% off, which can stretch your budget for inference, storage, and the rest of the stack.
Tips for Getting the Most Out of Your Credits
- Choose smaller models when you can, because neuron cost varies significantly by model size and you’ll get more calls per day.
- Watch your usage in the Cloudflare dashboard under Workers AI analytics, since it reports both neurons and conventional units like tokens, seconds, and images.
- Route requests through AI Gateway (free) for logging, caching identical prompts, and fallback behavior, which can save a surprising amount of compute.
- If you need customization, consider LoRA adapters on supported base models (Llama, Mistral, Gemma) rather than jumping straight to the biggest model.
- Build full-stack on Cloudflare’s free developer platform pieces (Workers, Vectorize, D1, R2, KV), so you’re not paying elsewhere while “saving” on inference.
Frequently Asked Questions
You get 10,000 Neurons per day, which Cloudflare estimates is roughly 1,300 LLM responses, about 12,500 embeddings, around 2,000 small image generations (512×512), or about 8,300 image classifications. Neuron cost varies by model, so bigger models will burn through the daily budget much faster. The good news is the allowance resets every day at 00:00 UTC, so you can treat it like a daily “compute stipend.” If you’re building a demo, split workloads: use smaller models for most requests and reserve the heavy one for the handful of cases that actually need it.
No.
There’s no time limit, but the 10,000 Neurons reset daily at 00:00 UTC.
Yes. If you have Cloudflare credits you won’t use before they expire, you can list them on AI Credit Mart and sell them at up to 70% of face value. Companies regularly list surplus credits from startup programs and enterprise agreements.
AI Credit Mart has discounted Cloudflare credits available from companies with surplus allocations. Prices are typically 30-70% below retail.
For Workers AI’s free tier, they don’t “expire”; the daily neuron quota simply resets at 00:00 UTC.
Yes, Workers AI supports OpenAI-compatible endpoints for chat completions, text completions, and embeddings, so you can swap the base URL and keep standard parameters like temperature and max_tokens.
Yes. Cloudflare notes that wrangler dev still consumes your Cloudflare account’s neuron quota, so local testing can quietly burn your daily allowance.
10,000 free Neurons a day is enough to ship a real prototype, and the “no credit card, no time limit” part is honestly rare. Use the reset to your advantage, and if you end up needing more Cloudflare credits later, you’ve got options.
Your AI credits are losing value every day
Join the marketplace and start trading unused credits today.