Cloudflare debuts Clef-omni, an open-weight decision model supporting audio and video input alongside text and image, and cuts Clef-flash's price below Jev's
Following last week's release of Clef and Clef-flash, Cloudflare's open-weight decision models, we decided to bring forth more gifts.
Context & Ripple Effects
Cloudflare introduced Clef and Clef-flash on October 2 as open-weight multimodal decision models positioned against Jev. Clef-omni extends that initial Clef model rollout rather than marking Cloudflare's first move into the category.
Jev had quickly become a prominent reference point for AI-tool selection: TypeSafe CEO Diogo Almeida said it was processing about one trillion tokens a day and used by roughly a quarter of Fortune 500 companies. That adoption makes Cloudflare's price comparison a bid for workloads already evaluating Jev.
First-order effects
- Cloudflare gives developers a Clef option that accepts audio and video as well as text and images, while lowering Clef-flash's stated price below Jev's.
- Teams using Cloudflare's models for PII discovery, malicious-domain detection, and documentation-spam triage, as described by a Cloudflare employee, can evaluate those tasks with an expanded input set.
Second-order effects
- Jev faces a more direct comparison on both multimodal capability and inference cost, increasing pressure to defend its value for tool-selection workloads.
- Cloudflare can use low-priced Clef-flash as an entry point for developers, with Clef-omni broadening the kinds of media-heavy workloads those customers can place on its model stack.
Third-order effects
- If comparable open-weight models continue to add modalities while cutting inference prices, model competition will shift toward the operational layers around them—deployment, routing, and application integration—rather than model access alone.
- The move reinforces open-weight commercialization: providers can distribute model weights while competing for recurring usage through hosted inference and adjacent cloud services.
The trend: Open-weight AI providers are pairing broader multimodal inputs with lower inference prices to compete for production workloads against fast-growing model platforms.