Qwen3.5-Omni-Plus
qwen-qwen3-5-omni-plus
fields
| nameREQ | Qwen3.5-Omni-Plus | awesomeagents.ai | 2026-05-04 |
| vendorREQ | Alibaba Qwen | awesomeagents.ai | 2026-05-04 |
| released_at | 2026-03-30 | awesomeagents.ai | 2026-05-04 |
| modality | text, image, audio, video | awesomeagents.ai | 2026-05-04 |
| access | closed | awesomeagents.ai | 2026-05-04 |
| context_window | 256000 | awesomeagents.ai | 2026-05-04 |
| parameters | ~30B total / ~3B active | awesomeagents.ai | 2026-05-04 |
| benchmark_score | 82.2 | awesomeagents.ai | 2026-05-04 |
| benchmark_name | MMAU | awesomeagents.ai | 2026-05-04 |
| announcement_url | https://qwen.ai/blog?id=qwen3.5-omni | awesomeagents.ai | 2026-05-04 |
history · 10 fields · 10 revisions
name1 revision
Qwen3.5-Omni-Plus
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.
vendor1 revision
Alibaba Qwen
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.
released_at1 revision
2026-03-30
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.
modality1 revision
text, image, audio, video
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.
access1 revision
closed
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.
context_window1 revision
256000
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.
parameters1 revision
~30B total / ~3B active
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.
benchmark_score1 revision
82.2
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.
benchmark_name1 revision
MMAU
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.
announcement_url1 revision
https://qwen.ai/blog?id=qwen3.5-omni
current
Qwen3.5-Omni-Plus released March 30, 2026. Thinker-Talker architecture with Hybrid-Attention MoE. Takes text, image, audio, video as input and produces text plus streaming speech. 256K context, 113-language speech recognition.