Our take
Cloudflare Workers AI runs downloadable models across the Cloudflare global network, priced for application developers already on the platform. It is best thought of as platform-integrated inference rather than a standalone model host.
Pick this for applications already on Cloudflare Workers or Pages that want zero-integration inference, or for edge latency with inference near users without managing regions. Use it if your team is GDPR-conscious and values verified EU and SOC 2 credentials. Skip it if you need the lowest price on flagship downloadable models or a very large model catalogue.
- Inference is co-located with Workers, KV and R2; no separate vendor.
- Verified EU endpoint, SOC 2 and no training on prompts.
- Mid-pack pricing on flagship downloadable models compared with the cheapest hosts.
- Curated catalogue: 12 tracked offers versus 61 at DeepInfra.
Privacy & data handling
- ✓
- Yes
- ✗
- No
- No answer on record
The colour says whether the answer favours you, not whether it is a yes — not training on your prompts earns a green cross, no zero-retention option earns a red one.