As of 2026-09-26, Alibaba's Qwen team shipped Qwen3.8-Omni-Flash on 2026-09-18: native omni (text, image, audio, and video in; text out) with a 1M context window. If you need chapters from a meeting or a long video, paste the link — you do not have to wire an omni API.
Paste a long-form video or meeting URL — BibiGPT turns it into chapters, a transcript, and follow-up Q&A.
As of 2026-09-26, Alibaba's Qwen team shipped Qwen3.8-Omni-Flash on 2026-09-18: native omni (text/image/audio/video → text), a 1M context window, and a reported +26% lift versus Qwen3.5-Omni-Plus on 30 evals. Last updated 2026-09-26.
On 2026-09-18 the Qwen team shipped Qwen3.8-Omni-Flash as a native omni model: one stack that takes text, image, audio, and video and returns text. It is live on Qwen AI and Alibaba Cloud Model Studio. This is not the 2026-08-26 text Flash-Next checkpoint.
Qwen describes a single model that accepts text, image, audio, and video and writes text. That is a different product from Qwen3.8-Flash (text multimodal MoE) and from Qwen3-ASR (speech recognition only).
The documented context window is one million tokens, with a 131K max output. A long meeting transcript can fit in one call. A wide window still does not file chapters or timestamps for you.
Qwen reports a +26% average lift versus Qwen3.5-Omni-Plus across 30 evals, and large drops on AliMeeting DER / cpWER (88.11 / 89.61 → 3.35 / 17.18). Those are vendor scores, not a BibiGPT benchmark.
A cheaper omni API helps teams that already wire Alibaba Cloud. It does not, by itself, give a student timestamped notes from a two-hour Bilibili upload or a recorded standup. BibiGPT is the product for that second job.
Paste a link, get chapters, a transcript, and follow-up Q&A. BibiGPT supports Qwen3.8 Omni Flash for long-form notes — you do not pick an omni recipe.
Audio and video in is useful. A 90-minute meeting still needs timestamps and an export. One cheap API call that you never file is wasted; a chaptered summary is a document you can search next week.
Qwen3.8-Flash is the August text / multimodal MoE. Gemini 3.8 Flash is Google's same-week competitor on AV tasks. Qwen3-ASR is speech recognition. This page is the 2026-09-18 Omni-Flash event. One intent, one URL.
Headline facts from Qwen's 2026-09-18 Qwen3.8-Omni-Flash release.
Qwen released Qwen3.8-Omni-Flash the same day on Qwen AI and Alibaba Cloud Model Studio (DashScope id qwen3.8-omni-flash). It is a hosted omni SKU, not an open-weight Flash-Next drop.
One model accepts four input types and writes text. That is a different job from Qwen3.8-Flash (text multimodal MoE) and from Qwen3-ASR (speech recognition).
Qwen documents a one-million-token window. LiteLLM's day-0 note lists 131K max output. Long meetings fit in one call; they still need chapters after the call.
Qwen reports +26% average on 30 evals versus Qwen3.5-Omni-Plus, AliMeeting DER/cpWER 88.11/89.61 → 3.35/17.18, and AV scores close to Gemini 3.8 Flash. Vendor numbers, restated here.
LiteLLM listed $0.15 input, $0.016 cached, $0.47 output per million tokens, and video input about 89% cheaper than Qwen3.5-Omni-Plus. Confirm on Model Studio before you budget.
Where a native omni Qwen SKU matters — and where a summarizer is the actual product.
Use Qwen3.8-Omni-Flash in your own stack for omni understanding. Then still paste the finished lecture or standup into BibiGPT so humans get chapters instead of a raw token dump.
A two-hour Bilibili upload or a recorded standup does not need an omni API brief. Paste the URL, export the transcript, keep asking questions. That is the BibiGPT path.
Qwen3.8-Flash is the August text MoE. Qwen3-ASR is speech recognition. Gemini 3.8 Flash is the Google AV competitor. This page is Omni-Flash only. One model-family event, one URL.
Meeting and long-video notes workflows that pair with this release.
Paste a link and get chapters, key points, and a clean summary.
Long B-site videos into structured notes.
Turn a podcast episode into searchable notes.
The August text / multimodal Flash SKU on its own URL.
Qwen's speech-recognition family — a different product.
Google's same-period AV Flash competitor.
Specs on this page come from Qwen's 2026-09-18 announcement and LiteLLM's same-week day-0 note.
Qwen shipped Qwen3.8-Omni-Flash on 2026-09-18 as a native omni model (text/image/audio/video → text) with a 1M context window, live on Qwen AI and Alibaba Cloud Model Studio. Qwen reports +26% versus Qwen3.5-Omni-Plus on 30 evals and AliMeeting DER/cpWER 88.11/89.61 → 3.35/17.18.
Qwen — Qwen3.8-Omni-Flash ↗LiteLLM's 2026-09-18 day-0 note lists DashScope id qwen3.8-omni-flash, 1M context, 131K max output, international list prices of $0.15 input / $0.016 cached / $0.47 output per million tokens, and video input about 89% cheaper than Qwen3.5-Omni-Plus.
LiteLLM — Qwen3.8-Omni-Flash day-0 ↗A native omni model takes more than one media type in the same stack — here text, image, audio, and video — and writes a single text reply. It is not a cascade of a separate ASR model plus a text LLM. Qwen3.8-Omni-Flash is documented as that kind of stack.
A one-million-token window can hold a long transcript plus instructions in one call. It does not, by itself, produce chapters, timestamps, or an export you can search next week. Those artifacts are the job of a notes product.
Why people use BibiGPT to turn videos into text every day.
Trusted by 50,000+ users worldwide
“I paste a link and get clean captions in seconds — it saves me hours of retyping every single week.”
Maya R.
Content Creator · Repurposes short videos
“Exporting the transcript lets me review new words at my own pace instead of pausing the video constantly.”
Daniel K.
Language Learner · Studies with real videos
“Accurate, timestamped text I can quote directly. It has quietly become part of my daily workflow.”
Priya S.
Researcher · Cites public talks
FAQ'S
Ask us anything!
1 Install a video-summary skill into DeepSeek Harness in one copy-paste — SKILL.md matches Claude Code. Or skip dsh: paste a Bilibili or YouTube link in the browser and get a timestamped summary.
2 Best Bilibili AI video summary tool 2026? Compare 5 options, then paste a Bilibili link for a free AI recap, mind map, and transcript-style takeaways — no login needed.
3 Extract Bilibili subtitles free: paste a BV/av link for a transcript even with no captions. Compare 5 tools for SRT export and AI summary. No signup to start.
Paste a YouTube, Bilibili, podcast, or meeting link. BibiGPT returns chapters, a transcript, and follow-up Q&A. Last updated 2026-09-26.
