← Back to Voice models
Alibaba (Qwen)Notable
Qwen3.5-Omni Realtime
Release date unpublished
The realtime audio and video conversation variants of Alibaba's Qwen3.5-Omni, in Plus and Flash tiers, on Model Studio.
- Price per minute
- —
- Pricing
- Singapore, per 1M tokens. Plus Realtime: input text/image/video $2.10, audio $16.50; output text $12.40, audio $62.00. Flash Realtime: input $0.55 / audio $4.50; output text $3.30, audio $17.70.
- Inputs
- audio, text, image, video
- Tool calling
- Yes
- Languages
- —
- Voices
- —
- Big Bench Audio
- 98.7%
- Access
- Generally available
- Licence
- Proprietary
Strengths & weaknesses
Strengths
- Big Bench Audio 98.7% on Plus, #5 on AA
- Text, audio, image and video input
- Semantic VAD for interruptions, plus function calling and web search
Weaknesses
- Plus audio output at $62 per 1M tokens
- Flash scores only 59.0% on Big Bench Audio (AA)
- Superseded for realtime audio and video by Qwen3.8-Omni-Flash-Realtime
Notable: #5 on AA Big Bench Audio (Qwen3.5 Omni Plus Realtime, 98.7%); Flash Realtime is #35 at 59.0%. releaseDate is null: the dated snapshot ID is qwen3.5-omni-plus-realtime-2026-03-15, and secondary sources date the Qwen3.5-Omni launch to about 2026-03-30; the qwen.ai blog didn't render. 113 input languages/dialects; 29 output languages + 7 dialects (Model Studio).
Links
API model strings
- alibaba
qwen3.5-omni-plus-realtime - alibaba
qwen3.5-omni-flash-realtime