Workers AI
An edge AI inference platform from Cloudflare that enables developers to invoke pre-built AI models, including the Whisper speech recognition model, at global edge nodes. Usage is measured in Neurons, with free accounts receiving 10,000 Neurons per day.
Timeline (last 90 days)
Cloudflare's GLM-5.3 Workers AI endpoint was reported to have returned other customers' data in some long-prompt responses.
A discussion thread about the GLM-5.3 Workers AI model appeared on Hacker News, titled "Cross-request KV cache exposure", receiving 155 replies, 15 points and 5 comments.
According to the analysis in the post, as a relatively new model version, GLM-5.3's serving stack may not yet be fully polished for edge scenarios.
On edge inference platforms, request scheduling is highly distributed, instance lifecycles are short, and scaling is dynamic, so cache state is difficult to audit uniformly and forcibly clean up the way it is in centralized data centers.
According to developer analysis in the Hacker News discussion, GLM-5.3 in Workers AI may have caused different requests to share cache content that should not have been shared, due to cache key design or gaps in request isolation.
Workers AI is Cloudflare's edge inference platform, focused on low latency and tight coupling with the Workers runtime.
OpenRouter's catalog currently lists only text and image inputs for clef-omni; if users need audio or video, they should go directly through Workers AI.
The Workers AI documentation notes audio at about 780 tokens per minute, and video at up to about 15,400 tokens per minute at the highest resolution (about 8,600 at 480p); at $0.15 per million, one minute of highest-resolution video with sound comes to about a quarter of a cent.
clef-omni's hosted ID on Workers AI is @cf/cloudflare/clef-omni, priced at $0.15 per million input tokens, with output free.
Cloudflare's Workers AI platform uses Neurons as the unit of measurement for AI inference consumption, and free accounts receive an allowance of 10,000 Neurons per day.
1 more timeline events
All Facts (8)
On edge inference platforms, request scheduling is highly distributed, instance lifecycles are short, and scaling is dynamic, so cache state is difficult to audit uniformly and forcibly clean up the way it is in centralized data centers.
50%UnverifiedAccording to the analysis in the post, as a relatively new model version, GLM-5.3's serving stack may not yet be fully polished for edge scenarios.
50%UnverifiedAccording to developer analysis in the Hacker News discussion, GLM-5.3 in Workers AI may have caused different requests to share cache content that should not have been shared, due to cache key design or gaps in request isolation.
50%UnverifiedWorkers AI is Cloudflare's edge inference platform, focused on low latency and tight coupling with the Workers runtime.
50%UnverifiedCloudflare's Workers AI platform uses Neurons as the unit of measurement for AI inference consumption, and free accounts receive an allowance of 10,000 Neurons per day.
50%UnverifiedAn AI agent running in the Cloudflare Workers environment cannot run large models locally on edge nodes and must call external APIs such as OpenAI, Anthropic, or Cloudflare's own Workers AI.
50%UnverifiedCloudflare's GLM-5.3 Workers AI endpoint was reported to have returned other customers' data in some long-prompt responses.
50%UnverifiedA discussion thread about the GLM-5.3 Workers AI model appeared on Hacker News, titled "Cross-request KV cache exposure", receiving 155 replies, 15 points and 5 comments.
50%