See how ChatGPT, Perplexity and Google AI Overviews describe you today
Free AI Visibility AuditQwen3.8 Omni Flash is an omni-modal reasoning model from Alibaba's Qwen family that handles text, image, audio and video in a single model, and is the first Qwen release built around agentic capabilities with native audio-video understanding.
QwenVerified 25/09/2026Released 21/09/202601 / decision
Each figure sits next to the middle of the field, so it can be read as dear or cheap, wide or narrow, rather than floating on its own. Compared against the 231 models we track, not against an absolute standard.
Capability
Not measured
No published benchmark scores yet
Input / 1M tokens
$0.15
field median $0.43
Output / 1M tokens
$0.47
field median $1.81
Context
1,000K
field median 500K
02 / overview
Four questions, answered separately, because somebody arrives at one of them rather than at the top of the page.
The job it was built for, and the job it is not for.
Qwen3.8 Omni Flash is a reasoning model that takes text, images, audio and video as input rather than text alone, with agentic use as the design target. Alibaba positions it for audio-video analysis and summarisation, so the job is watching or listening to something and producing a useful account of it. It is not a specialist text-only writing or coding model, and the omni-modal framing is the point rather than a side feature.
Where it sits in its line, and when a sibling is the better pick.
Qwen3.8 Omni Flash was first seen in September 2026, and it sits in the Qwen line as the Flash tier, the variant tuned for throughput and low cost rather than maximum depth. Pick it when the work involves audio or video that needs to be understood natively, or when an agent needs to read several modalities in one pass. If the task is plain text reasoning with no media in it, a text-focused model will usually be the cleaner choice.
The API, the apps it powers, and what its limits let you do.
Qwen3.8 Omni Flash is reached through the Qwen API from Alibaba, and takes text, image, audio and video in the same request. Its very large context window means long recordings, transcripts and accompanying documents can go in together rather than being chunked and stitched, and the generous output ceiling leaves room for full summaries and structured extractions rather than clipped answers.
What changes because this exists, or why it does not.
Qwen3.8 Omni Flash makes audio and video a first-class input instead of something you pre-process into a transcript and hope nothing important was lost. A team that previously ran speech-to-text, then vision, then a separate reasoning pass can collapse that into one call, and the agentic framing means the model is intended to act on what it has seen rather than just describe it.
Follows Qwen3.8 Max (0902). Superseded by Qwen3.8 Max Prime. See the whole line.
03 / evidence
Split by category, so a strong number never hides a thin evidence base. Verified means we read it on the benchmark's own published results; a provider's claim about its own model is shown and labelled rather than dropped.
No published benchmark scores for this model yet.
We publish a score only where we can link the result to where it was published. Until a benchmark result for this model exists in a source we read, this section stays empty rather than being filled with a provider's marketing figure.
How we decide what counts as evidence04 / ledger
Every published row, grouped by category, each compared with the best published score on the same benchmark. 'Is 64% good' is a question nobody can answer; '26 points behind the leader' is one anybody can.
Nothing in the ledger yet.
Each row here carries a score, the benchmark it came from, what the leading model scored on the same test, and a link to the published result. Rows appear as results are published and read.
How we decide what counts as evidence05 / capability
Where this model is strong, and against how many peers. Ranks are against models with evidence in that category, not against everything we track: ranking against models nobody tested would rank who published, not who is better.
No category scores to shape yet.
A category score is the weighted mean of the benchmarks published for it. With no published rows there is nothing to average, and an empty chart drawn at zero would say something false.
How we decide what counts as evidence06 / cost
List API rates as last read from the provider, with the source on every row, plus every change we have recorded since we started tracking it.
| Charge | Price | Unit | Read on | Source |
|---|---|---|---|---|
| Input | $0.15 | per 1M tokens | 2026-09-24 | Check |
| Output | $0.47 | per 1M tokens | 2026-09-24 | Check |
Qwen3.8 Omni Flash is priced at the cheap end of the field, which is what the Flash label is signalling, and output costs only a few times more than input rather than the wide gap seen on frontier tiers. For multimodal work at volume, where long video and audio contexts inflate token counts quickly, that difference is the main reason to choose it.
| Workload | Input / month | Output / month | Cost |
|---|---|---|---|
| A small product team | 20M tokens | 5M tokens | $5.35 |
| A busy support assistant | 200M tokens | 40M tokens | $48.80 |
| A document pipeline | 1000M tokens | 100M tokens | $197.00 |
List API rates, no caching and no batch discount, which both providers offer and which change the answer a great deal. Treat these as the ceiling, not the bill.
07 / specs
As listed by the provider's own catalogue and re-read every few hours. Anything absent is absent there too.
| Context window | 1,000,000 tokens |
|---|---|
| Maximum output | 131,072 tokens |
| Modalities | text, image, audio, video |
| Released | 21/09/2026 |
| Status | Current |
| Catalogue identifier | qwen/qwen3.8-omni-flash |
08 / lineage
What this model replaced, what replaced it, and what else its provider has in the field.
Also from Qwen
09 / line
Every model in this family in release order, so a page from eight months ago says in one glance that two newer ones exist.
Came before
← Qwen3.8 Max (0902)Came after
Qwen3.8 Max Prime →Ordered by release date and worked out from the naming, so a new member slots in as soon as its page exists. A retired model keeps its page and its place in the line.
10 / notes
What this model changes for a brand trying to be cited in AI answers, and every change we have logged since it launched.
What it changes for you
If your buyers use tools built on Qwen3.8 Omni Flash, some of what they learn about your category will come from a model that has watched a demo video or listened to a webinar, not just read your pages. That makes captions, transcripts and spoken claims part of what gets cited, so the things you say on camera should match the things you publish in text. It also means the cheap multimodal tier is now viable for everyday product research, so expect more of it.
Most buyers meet Qwen3.8 Omni Flash through the Qwen API rather than a consumer chat window, usually inside a product someone else has built on it. Expect to encounter it behind media analysis tools, meeting and call summarisers, and agent workflows that need to read a screen recording or a clip as part of a longer task.
Nothing published here yet. Changes appear within hours of a provider announcing them.
11 / questions
The things people ask about this model, answered from what is on this page rather than from anywhere else.
Models change what gets cited. We measure whether AI answers name your brand or your competitors across every assistant, and show you what to change.