Qwen Audio 3.0 Realtime Flash measured on 12 of 34 Realtime benchmarks12
BenchmarkByScoreRank
Agents4
Multimodal6
Legacy · saturated or no longer reported2
No benchmark matches
Ranking?
Full ranking →···
10GPT Realtime 1.585.111Qwen Audio 3.0 Realtime Flash81.712GPT Realtime Mini79.6Strong at
Alternatives?
ModelScoreCostPriceWhat changes
ModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
Benchmarks
in Speech & VoiceBig Bench Audio · AA v1.296.1#11
Full-Duplex-Bench · AA subset v1/v1.596.9#5
Time to first audio · AA1.6#23
AA Speech Agent Arena · overall · Elo847#16
AA Speech Agent Arena · overall · Task Success81.7#11
AA Speech Agent Arena · agentic · Elo847#15
AA Speech Agent Arena · non-agentic · Elo847#18
τ-Voice · AA · aggregate35.9#10
τ-Voice · AA · airline54#9
τ-Voice · AA · retail29.2#11
τ-Voice · AA · telecom24.6#8
FAQ
Is Qwen Audio 3.0 Realtime Flash free to use?›
There is no free API; the cheapest paid price is $0.93 per 1M input tokens.
How good is Qwen Audio 3.0 Realtime Flash, and what is it best at?›
#10 of 17 in Realtime.
Is Qwen Audio 3.0 Realtime Flash better than GPT Realtime 1.5?›
On AA Speech Agent Arena · overall · Task Success, Qwen Audio 3.0 Realtime Flash scores 81.7 vs 85.1 — about the same.
How can I use Qwen Audio 3.0 Realtime Flash?›
Through the Alibaba (Qwen) API.
When was Qwen Audio 3.0 Realtime Flash released? Is there a newer version?›
Released 2026-08-10. It is the current version in its line.
Is Qwen Audio 3.0 Realtime Flash open source?›
No, Qwen Audio 3.0 Realtime Flash is a closed model available through an API.
What is Qwen Audio 3.0 Realtime Flash?›
Qwen Audio 3.0 Realtime Flash is a closed speech & voice model from Alibaba (Qwen), released Aug 2026.