AI news · Cloudflare · models
Cloudflare adds audio and video to its Clef decision models
Clef-omni reads sound and video in a single call, and Cloudflare says a price cut puts Clef-flash below TypeSafe’s Jev per token.
Cloudflare released Clef-omni on Friday 9 October 2026, an open-weight decision model that accepts audio and video alongside text and images, the company announced. It also cut the price of Clef-flash and sped up its hosted Clef model, a week after launching the family.
A decision model returns a pick from a fixed set of options with a confidence score, instead of writing text. Most decision models so far read only text, Cloudflare says, a pattern dating back to TypeSafe’s Jev. Clef-omni is built on the Qwen3-Omni-30B-A3B mixture-of-experts model with low-rank adapters, takes WAV or MP3 audio and MP4 or WebM video, and scores a 21-second clip with sound in about 1.5 seconds, Cloudflare says. Its weights are on Hugging Face.
Hosted prices per million input tokens: Clef-flash falls from $0.09 to $0.038, Clef stays at $0.24 and Clef-omni starts at $0.15. The cheaper Clef-flash also gets a smaller hosted context window, down from 64,000 to 24,000 tokens, Runtimewire reports. On TypeSafe’s invoice-processing workflow, as run by Cloudflare, Clef-omni scores 60.2 against 61.8 for Jev and 64.7 for Clef.
Cloudflare’s pitch to teams that sort calls, clips or screenshots is one model call where a transcription step used to sit.
Sources
- Cloudflare: Introducing Clef-omni with full multimodality, plus a faster Clef and a cheaper Clef-flash. Checked
- Runtimewire: Cloudflare puts audio and video into its decision model with Clef-omni. Checked
Alexandre S.
Alexandre S. is a creative ad strategist and copywriter with ten years' experience in digital marketing across multiple channels.