10 Oct 2026
OllyGarden raises $4m seed to clean up OpenTelemetry data
Founded in 2025 by OpenTelemetry contributors, the company launched Minimum Viable Instrumentation with the round. Next Frontier Capital,…
Cloudflare's multimodal decision model, released 9 October 2026 under Apache 2.0: takes text, images, audio and video and returns a score for each allowed answer instead of free text. 30B mixture of experts with 3B active.
Not yet independently verified. Every detail here comes from Cloudflare's own blog post of 9 October 2026 and the model card on Hugging Face; no independent report was found. The accuracy figures on the card are from Cloudflare's own internal evaluation, so no score is stored. The blog post names no licence; the Apache 2.0 licence is taken from the model card. We will update this when it can be confirmed, and remove this note.
Clef-omni is an open-weight decision model from Cloudflare that takes audio, video, images and text and returns a probability for each answer option it is given. Cloudflare released it on 9 October 2026 as the newest member of its Clef family. It does not write text, so there is nothing to parse in its output.
The caller sends a state and a schema of typed questions, and the model scores the allowed options in a single forward pass. Question types are choice, yes or no, and score. The Hugging Face card says it was post-trained from Qwen3-Omni-30B-A3B-Instruct, a mixture-of-experts model, with a small schema head reading the backbone's final hidden states. Earlier Clef models handled images and video frames but not audio. Video is sampled at two frames per second.
Cloudflare gives median latencies of about 130 milliseconds for text, 150 for images and a few hundred for audio, with a 21-second video with sound taking about 1.5 seconds. These are the vendor's figures.
Cloudflare pitches it at teams that need classification or detection across several kinds of input without training a specialised model. Its examples include spam detection, phishing moderation, scanning for personal data and spotting malicious domains. Running the open weights takes about 64 GB of GPU memory in bfloat16.
On Cloudflare's own benchmarks it leads the family on three of ten tests, including BANKING77 at 94.8 macro-F1, but trails other Clef models on most others.
| Benchmark | Official | Community avg |
|---|---|---|
| No benchmark scores yet. Be the first to add one. | ||
“Official” values are editor-approved and feed the ranking. “Community avg” is the mean of member submissions (shown for transparency; it never affects the ranking until an editor approves a value).
Sign in to add a benchmark score for this model.
An earned signal from verification, reviews, awards, transparency and engagement - the vendor can't buy it.
Updated 10/10/2026
Used it? Your experience helps other buyers decide.
Write a reviewNo questions yet. Be the first to ask about Clef-omni.
Everything here links back to the same verified catalogue. Pick your next stop.