Qwen2.5-Omni
End-to-end omni-modal 7B: text, image, audio, and video in, text or speech out
About
Qwen2.5-Omni-7B, released March 26, 2025 under Apache 2.0, was the Qwen family's first end-to-end omni-modal model: one model takes text, images, audio, and video and answers in text or natural speech. Its Thinker-Talker architecture splits understanding from speech synthesis, with time-aligned positions keeping video and audio in sync for low-latency spoken replies. Alibaba reports state-of-the-art OmniBench results and a 4.51/5 speech quality score it describes as human-level. The 32K context and heavy VRAM needs (around 31GB for 15 seconds of video in BF16) are the real constraints; a minute of video needs a top-tier GPU.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
OmniBench SOTA, speech generation quality 4.51/5 (Alibaba-reported)
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Release 2025-03-26, Apache 2.0, Thinker-Talker architecture, 32,768 context verified via technical report and Alibaba coverage; OmniBench SOTA and 4.51/5 speech quality are Alibaba-reported; labeled as such; VRAM figures (31GB for 15s video BF16) from third-party analysis; memory-hungry caveat captured.