Accepts text, images, audio, and video with text-only output. Supports tool calling and a one-million-token context. Model Studio lists availability from September 17; the launch article followed September 20. Independent benchmark coverage is pending.
| Benchmark / source | Score |
|---|---|
GPQA Diamond Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 91% |
HLE Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; GPT-4o judge | 36.5% |
LiveCodeBench v6 Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 92.6% |
IFBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 81.5% |
DeepSWE 1.1 Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Best of Claude Code and mini-SWE-agent; 256K context; temperature 1.0, top_p 0.95 | 57.8% |
SWE-bench Pro Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Claude Code; corrected tasks; 256K context | 63.3% |
SWE-bench Multilingual Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; mini-SWE-agent; 256K context | 80.5% |
NL2Repo-Bench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Claude Code; repository retrieval commands disabled | 48.9% |
CoWorkBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; In-house productivity evaluation | 75.3% |
ClawEval-MM Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Pass@3 | 60.4% |
ClawEval-MM Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Mean across three trials | 61.9% |
AndroidWorld Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 87.1% |
Vision2Web Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Claude Code; GPT-5.4 judge; category average | 62.9% |
ERQA Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 71% |
LVBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 76.9% |
RealWorldQA Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 87.7% |
MathVision Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Without CI; corrected annotations | 91.8% |
MathVision Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; With CI; corrected annotations | 96.2% |
CharXiv RQ Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Without CI | 83.5% |
CharXiv RQ Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; With CI | 91.4% |
WildClawBench-MM Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Claude Code | 71% |
UniClawBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; OpenClaw | 69.6% |
AgenticVBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Claude Code | 36.8% |
OmniGAIA Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; No external agent harness | 74% |
OmniVideoBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Agentic understanding; context retained across turns | 67.8% |
OmniVideoBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Static understanding | 63.4% |
Video-MME-v2 Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Static | 65% |
Video-MME-v2 Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Qwen Code | 71.3% |
LVOmniBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Static | 63.3% |
LVOmniBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Qwen Code | 73.6% |
AliMeeting DER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 3.4% |
AliMeeting cpWER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 17.2% |
AISHELL-4 DER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 2.8% |
AISHELL-4 cpWER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 11.2% |
MagicData-RAMC DER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 5.7% |
MagicData-RAMC cpWER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 14.1% |
MLC-SLM English DER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 4% |
MLC-SLM English cpWER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 14.2% |
WenetSpeech Net WER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 4.8% |
WenetSpeech Meeting WER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; Lower is better | 4.6% |
FLEURS-ASR WER Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; 60 languages; lower is better | 9.3% |
FLEURS-S2TT BLEU Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses; 60 languages; higher is better | 31.8 score |
SpotSoundBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 67.2% |
MMAU Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 81.8% |
MMAR Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 79.8% |
MMSU Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 82.1% |
MuchoMusic-RUL Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 72.6% |
HumMusQA Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 75.8% |
MusTBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 50.6% |
Audio MultiChallenge Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 71.5% |
WildSpeech Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 74.3% |
VoiceBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 91.6% |
DailyOmni Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 85.1% |
WorldSense Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 68.5% |
AVUT Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 86.6% |
JoinAVBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 75.9% |
AVSpeaker Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 77.2% |
MMOU Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 78.8% |
OmniCloze Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 63.2% |
OmniCap-IF CSR Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 80.6% |
OmniCap-IF ISR Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 28.2% |
QIVD Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 69.6% |
OmniVChat-Bench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 89% |
ProactiveVideoQA Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 61.5% |
StreamingBench Provider reportQwen technical report v1, Tables 2-7; provider evaluation with benchmark-specific harnesses | 80.8% |