Introducing Clef-omni with full multimodality, plus a faster Clef and a cheaper Clef-flash
We are expanding the Clef decision model family with Clef-omni, natively processing audio, video, images, and text in a single pipeline. We’ve also lowered Clef-flash pricing and boosted Clef inference speeds by up to 2.0x.
Cloudflare introduces Clef-omni, a multimodal AI decision model supporting audio, video, images, and text, alongside performance and pricing improvements for Clef and Clef-flash.
What happened
Cloudflare expanded its Clef decision model family with Clef-omni, a multimodal model handling audio, video, images, and text. They also reduced Clef-flash pricing, improved Clef's inference speed, and provided technical details about model architecture and benchmarks.
Why it matters
Multimodal support simplifies AI integration for complex workflows, while cost reductions and speed improvements make AI more accessible for developers and businesses. The technical advancements enable efficient processing of diverse data types.
- Multimodal models simplify complex AI workflows by handling diverse data types in a single pipeline.
- Cost reductions and speed improvements make AI more accessible for enterprise adoption.
- Technical innovations like MoE architecture and attention routing enhance model efficiency and accuracy.
The bigger picture
Enables more versatile AI applications across industries, reduces operational costs for businesses, and accelerates real-time decision-making processes through improved speed and multimodal capabilities.
Cloudflare's Clef-omni release marks a significant advancement in multimodal AI, offering improved performance, affordability, and versatility for developers and businesses.
Technical details
- Built on Qwen3-Omni-30B-A3B-Instruct foundation with mixture-of-experts (MoE) architecture
- Two-stage attention routing for evidence gathering and confidence scoring
- Low-rank adapters (LoRA) for training efficiency
- Schema-constrained scoring with lexical grammar preservation
- SGLang optimization for serving infrastructure
FAQ
What makes Clef-omni unique compared to previous models?
Clef-omni is the first Cloudflare model to natively process audio, video, images, and text in a single pipeline, eliminating the need for separate processing steps.
How does the pricing change affect users?
Clef-flash is now 58% cheaper than before, making it more accessible for cost-sensitive workflows, though with a reduced context window of 24k tokens.
What benchmarks were used to evaluate performance?
The models were tested on industry benchmarks like BFCL, ToolRet, API-Bank, and domain-specific tasks, showing strong results across multiple metrics.
This summary and analysis were generated with AI from the original article and may contain errors. Read the original for the full story.
Read the full story
Originally published by Moltworker Blog on Oct 9, 2026.
Comments
Sign in to join the discussion.