Cloudflare launches Clef-omni, cuts Clef-flash price by 58%

Share
Cloudflare launches Clef-omni, cuts Clef-flash price by 58%

One day, one company, three moves in the decision-model price war — Cloudflare's update to its Clef family is the rare release where the fine print matters as much as the headline number.

Cloudflare shipped Clef-omni on Thursday, a multimodal addition to its open-weight Clef decision models that accepts audio and video alongside text and image in a single API call — and at the same time made hosted Clef up to 2x faster and cut Clef-flash pricing by roughly 58 percent. Clef-omni takes WAV or MP3 audio and MP4 or WebM video; Cloudflare reports median latencies around 130 milliseconds for text, 150 for images, and about 1.5 seconds for a 21-second video clip. It launches at $0.15 per million input tokens, sitting between Clef-flash and the full Clef at $0.24. Under the hood it is a 30B-A3B mixture-of-experts model post-trained from Alibaba's Qwen3-Omni-30B-A3B-Instruct — the same pattern Cloudflare used for the original Clef line, lean enough to run on Workers AI rather than a frontier-lab budget.

The speed gains come from the serving layer, not new weights: Cloudflare moved Clef onto the SGLang inference stack, cutting median latency for a typical 3,400-token request from 616 to 305 milliseconds and its p95 from 777 to 531 — the "up to 2x" in the announcement. Clef-flash input pricing drops from $0.09 to $0.038 per million tokens. The catch is the one worth reading: the hosted Clef-flash context window shrinks from 64k to 24k tokens. Cloudflare says only 0.24 percent of requests ever exceeded 24k, and self-hosted weights still carry the full 256k — but the shape of the deal is telling. In this market, context is the currency labs trade away to buy a lower price, and eight days after launch Cloudflare decided cheap-and-short beats slightly-roomier-and-pricier.

That framing lands harder with TypeSafe's Jev in the mirror: Jev's $870 million round at a $7.5 billion valuation this week shows investors will pay for the category Cloudflare is undercutting. We covered the original launch in Cloudflare open-sources the Clef models, aiming straight at Jev. One caveat before anyone builds on the numbers: the benchmarks in the announcement are Cloudflare's own — including its losses to Clef and Jev on TypeSafe's invoice-workflow evals — and RuntimeWire notes none have been independently audited.

What to watch: whether rivals match the $0.038 floor within weeks, and whether the 24k context trade becomes the industry's standard move or a Cloudflare-specific one.

If price and context window are now a dial rather than a fixed spec, which one would you give up first? Write to us via our contact page.

Get SiliconNoon in your inbox

One email a day with the AI stories that matter — what happened, the numbers, and why it matters. No spam, unsubscribe in one click.

Read more

Two fixed opening tokens push a base model past its RL version

Two fixed opening tokens push a base model past its RL version

A new paper from MIT, UC Berkeley, Washington University and the Allen Institute for AI says much of what reinforcement learning teaches a model may come down to how it starts an answer. One good prefix, fixed in advance, is enough to close the gap. Fix the first two tokens and Olmo-3-7B beats its own RL-trained twin. The paper — "Base Models Can Reason By Taking a Cue From Training Data" — tested what happens when researchers pre-fill a model's response opening instead of letting it choose.

Byteification turns Qwen3 and Llama 3 into byte-level models

Byteification turns Qwen3 and Llama 3 into byte-level models

The tokenizer is the last piece of the language-model pipeline that was never learned end to end — and a Nature paper out this week shows it can be swapped out for less than one percent of a pretraining budget. Researchers have retrofitted four open subword language models into true byte-level models using less than 1% of a typical pretraining budget, and the result — published 7 October in Nature — closes the performance gap that has kept byte-level models a niche curiosity for nearly a…

Jev 1.14.0 can read images — if you open its 'Danger Zone'

Jev 1.14.0 can read images — if you open its 'Danger Zone'

TypeSafe's decision model learned to look at pictures this morning, but the company shipped the capability behind a warning label instead of a launch announcement. Jev 1.14.0 adds image input, a new question type called Box, and a 262,000-token context window — all three shipped as experimental features behind a program TypeSafe calls the "Danger Zone." The features are off by default and have to be enabled per request through the API; TypeSafe's own documentation says they do not meet…

Kling 4.0 goes live with 30-second native video at 10-bit HDR

Kling 4.0 goes live with 30-second native video at 10-bit HDR

Kuaishou's Kling AI has spent two years becoming the video model people actually finish projects with. Its new flagship, Kling 4.0, went fully live on October 15 — and this round is less about prettier frames than about the two things that make a generated shot usable in a real edit: length and control. Kling 4.0 generates up to 30 seconds in a single native pass, outputs 10-bit HDR at 4K, 1080p and 21:9, and takes up to 15 reference assets in one task.