Alibaba CloudAudio modelAPI only

Qwen Audio 3.1 Realtime Flash

Summary

Qwen Audio 3.1 Realtime Flash is Alibaba's API model for real-time duplex speech dialogue. It accepts speech and text and can return speech and text. WebSocket, WebRTC, and AOQ are documented transport options. The model also supports function calling, web search, voice cloning, and several turn-taking modes.

The Model Studio documentation lists a 262,144-token context, up to 50 audio turns, and 300 seconds of cumulative audio in retained history. The Qwen-Audio-3.1-Realtime research paper publishes many evaluations for a system named Qwen-Audio-3.1-Realtime, but does not identify that system as the Flash API variant. Those scores are therefore not attributed here to Flash.

All specifications

Architecture

Model class: Proprietary real-time speech model; Source: Model class

Input and output

Input modalities: Audio and text; Source: Input modalities
Output modalities: Audio and text; audio output may include accompanying text output; Source: Output modalities

Licence and access

Access: Alibaba Cloud Model Studio API; Source: Access

API

API model ID: qwen-audio-3.1-realtime-flash; Source: API model ID

Further details

Interaction modes: server_vad, smart_turn, smart_turn_v2, and push-to-talk; Source: Interaction modes
Tools: Function calling and web search are supported and can be enabled together according to the current documentation; Source: Tools
Voice cloning: Supported; default voice is longanqian_v3.1; Source: Voice cloning
Transport protocols: WebSocket, WebRTC, and AOQ; Source: Transport protocols
Context limits: 262,144-token context, 245,760 input tokens, 16,384 output tokens, up to 50 audio turns, and 300 seconds of cumulative audio history; Source: Context limits

Prices

Singapore API pricing: Per 1M tokens: text input $0.23, audio input $0.93, text output $0.70, audio output $1.87. When audio and text are returned together, only audio output is billed.Source
Beijing API pricing: Per 1M tokens: text input $0.212, audio input $0.848, text output $0.636, audio output $1.696. When audio and text are returned together, only audio output is billed.Source

Sources and data date

Type	Evidence and data date
Research status	Research date October 29, 2026. 59 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Unresolved	Benchmarks for the exact Flash API variant Note The reviewed API documentation publishes no scores for this exact Flash variant. The research paper does not unambiguously map its evaluated version to the Flash API. No benchmark can therefore be safely attributed to this model at this time.
Not found in the inspected sources	Vals rows in captured payloads Note Nine complete Vals payloads dated 28 October were checked; no exact model ID or unambiguous model label. Limited to these saved benchmarks.
Unresolved	Uncaptured Arena Code/Speech leaderboards Note Text, Agent including Code/Chat/Work, and twelve WebDev views were captured. A separate general Code and Arena Speech leaderboard were not saved as complete tables.
Unresolved	Artificial Analysis catalog reconciliation Note The full current AA model and benchmark catalog was not re-reconciled for this profile in this sidecar. This sidecar makes no AA absence claim; native AA imports remain authoritative.
Not found in the inspected sources	VoiceArena STT exact profile row Note The saved VoiceArena speech-to-text results do not contain this exact model identity. This is limited to that capture and does not rule out future rows or other speech evaluators.
Additional source	Alibaba Cloud Qwen Audio 3.1 Realtime Flash documentation (retrieved October 29, 2026) · Alibaba Cloud Qwen Audio 3.1 Realtime Flash documentation · Editorial description reviewed October 29, 2026
Additional source	Qwen-Audio-3.1-Realtime paper, not mapped to Flash (retrieved October 29, 2026) · Qwen-Audio-3.1-Realtime research paper · Editorial description reviewed October 29, 2026
Additional source	Model metadata (retrieved October 29, 2026)1 2