Qwen

Qwen: Qwen3.8 Omni Flash

qwen/qwen3.8-omni-flash

canonical: qwen/qwen3.8-omni-flash-20260918

Audio inside a video clip

Hears clip audioprobed 2026-10-22

Probed: repeated a codeword that was only ever spoken on the audio track.

What it answered

SHAPE: yellow square
AUDIO: purple elephant

Descriptionfrom OpenRouter

Qwen3.8 Omni Flash is an omni-modal reasoning model from Alibaba, the first Qwen model built around agentic capabilities with native audio-video understanding. It is suited for audio-video analysis and summarization,...

Context length

1,000,000 tok

Max completion

131,072 tok

Knowledge cutoff

—

Expires

—

Added to OpenRouter

9/21/2026

Cache last refreshed

11/12/2026

Primary modality

text+image+audio+video->text

Source

OpenRouter

Input modalities

textimageaudiovideo

Output modalities

text

Pricing

Field	Per token (USD)	Per million
prompt	0.00000015	$0.150
completion	0.00000047	$0.470
input_cache_read	0.000000016	$0.016

Supported parameters(16)

frequency_penaltyinclude_reasoninglogprobsmax_tokenspresence_penaltyreasoningresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p

OpenRouter page

Rounds

2 answersin this database. A round opens every model's answer to the same prompt.

Oct 17, 2026Vision & PerceptionSeconds Until Out of Framein a video round answer
Oct 17, 2026Vision & PerceptionCount the Park Catsin a video round answer