Unverified50% confidenceFactExact time
clef-omni is based on the Qwen3-Omni-30B-A3B-Instruct base (using only the thinker, without speech output), with a total of 30B parameters, 3B activated, supporting text, JSON, images, audio, and video (sampled at 2fps), with a context of 64,000 tokens.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedClef-omni is capable of scoring and judging content across four modalities—text, images, audio, and video—in a single call.70% similarUnverifiedMiniMax Code CLI can connect to image, audio, and video tools, supporting multimodal asset processing.63% similarUnverifiedQwen3-Omni-Flash supports audio, images, and video on the input side, and can generate both text and speech modalities on the output side.62% similarVerifiedQwen3.6-27B supports multimodal capabilities, including text reasoning, image understanding, and video processing.62% similarUnverifiedBoth the Clef model and the Laya model with a Qwen3.5 backbone can be exported to GGUF format and served via llama.cpp, including multimodal decision models capable of understanding images.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1011197API
curl https://kongchang.com/api/v1/knowledge/claims/1011197MCP
get_claim(id=1011197)