Alibaba's Qwen3.8-Omni-Flash takes text, image, audio and video in one model, claims Gemini 3.8 Flash parity

Alibaba's Qwen3.8-Omni-Flash takes text, image, audio and video in one model, claims Gemini 3.8 Flash parity

Alibaba's Qwen team shipped Qwen3.8-Omni-Flash on September 14, 2026: one model that accepts text, image, audio and video in the same API request, carries a 1M-token context window, and reports a 25%+ average gain over Qwen3.5-Omni-Plus across 29 evaluations.
The model is not open weight. It ships only through Alibaba Cloud's DashScope API, with no downloadable checkpoint found on Hugging Face at publication time, unlike other Qwen and DeepSeek releases.
API pricing for audio and audio-visual input drops 93 to 98 percent, and a separate Realtime variant streams audio and video in and synthesized speech out over WebSocket or WebRTC for live sessions.
On OmniVideoBench, agentic mode scores 67.8 accuracy at 79,117 tokens per query versus 63.4 accuracy at 145,736 tokens for static mode, a 45.7 percent token cut with higher accuracy. Gemini 3.8 Flash still leads in absolute terms at 65.2 static and 70.1 agentic.
Speech support covers 74 recognition languages plus 39 Chinese dialects, and generation covers 29 languages plus 7 dialects, with OpenAI-compatible Chat Completions and Responses APIs.

AI video model releases

Alibaba's Qwen3.8-Omni-Flash takes text, image, audio and video in one model, claims Gemini 3.8 Flash parity Alibaba confirms Wan 3.0 launch for Aug 24, lands #2 on OpenArt Arena behind ByteDance's Seedance 2.5 Alibaba releases Qwen3.8-Omni-Flash, a hosted 1M-token omni model that plans what to watch in video HiDream launches HiDream-O1-Video-1.0, an omnimodal video model that debuts at No. 4 on Artificial Analysis image-to-video leaderboard HiDream launches HiDream V1 omnimodal video model, debuts at No. 4 on Artificial Analysis image-to-video leaderboard ShengShu Technology's Vidu S2 splits into a real-time avatar model and a real-time video editing model, at 720p output HiDream launches HiDream-O1-Video-1.0, debuts at No. 4 on Artificial Analysis image-to-video leaderboard with audio HiDream launches HiDream-O1-Video-1.0, an omnimodal 1080p video model debuting at No. 4 on Artificial Analysis Fal's H3 Max claims 35x throughput of MiniMax's H3 endpoint, tops two image-to-video leaderboards Alibaba's Qwen3.8-Omni-Flash cuts video input costs about 89% versus Qwen3.5-Omni-Plus Alibaba releases Qwen3.8-Omni-Flash with 1M-token context, claiming Gemini 3.8 Flash-level audio and video HiDream launches HiDream-O1-Video-1.0, an omnimodal video model debuting at No. 4 on Artificial Analysis Alibaba Cloud's Wan3.0 video model outputs a single 30-second long take and accepts up to five video references ShengShu Technology Unveils Vidu S2, Splitting Real-Time AI Video Into an Interactive Avatar Model and a Live Stream Editor HiDream launches HiDream-O1-Video-1.0, a native omnimodal video model debuting at No. 4 on Artificial Analysis Kling 3.0 launches with native 4K 60fps video, multi-shot storyboarding, and 5-language lip sync ShengShu unveils Vidu S2, adding real-time avatar interaction and live stream editing Shengshu Technology Launches Vidu S2 With Real-Time Editing, Dynamic Reference Images and VR Headset Streaming ByteDance Open-Sources Bernini, an Apache-2.0 Video Editor That Plans the Edit Before It Renders ByteDance Restricts Seedance 2.0 Video Model Three Days After Release, Following Viral Deepfake Demo Kuaishou launches Kling 3.0 with native 4K 60fps, 6-shot AI Director, and 5-language lip sync JD.com open-sources EchoWM interactive audio-visual world model and shows JoyAI-Echo1.5 clearing 10 minutes of continuous generation Kuaishou's Kling AI launches Kling 3.0 with native 4K 60fps video, multi-shot storyboarding, motion control and 5-language lip sync Black Forest Labs releases FLUX Video Edit, an API-only video editing model at $0.03 per second Kling AI Ships Kling 3.0: Native 4K 60fps, Multi-Shot Storyboarding, 5-Language Lip Sync Kuaishou Says Kling 3.0 Generates 4K 60fps Video With Six-Shot Sequences OpenVDN Releases VDN-H3 Video Patch Reporting 14.4-Second 768p Clips in 11.23 Seconds Kuaishou Announces Kling 3.0 With Claimed Native 4K 60fps and Six-Shot Video Generation Twelve Labs' Marengo becomes Amazon Bedrock's first video model JD.com Open-Sources JoyAI-EchoWM for Interactive Video and Audio World Generation fal releases H3 Max, claiming top image-to-video rankings and 3-second generation for 5-second clips Overworld releases Waypoint-1.5 real-time world model with 720p 60 FPS and 360p tiers Adobe puts Firefly, Veo, Runway, Kling and Luma video generation inside Premiere timelines NVIDIA Releases Sol-H3 Inference Code, Running MiniMax H3 Video Faster Than Playback on Eight B300 GPUs NVIDIA and SANA Release Sol-H3 Stack That Generates 5-Second MiniMax-H3 Video in 1.653 Seconds CreateAI open-sources Ruyi-Mini-7B image-to-video model for consumer GPUs World Labs says Atlas generates camera-controlled views and 3D scenes from sparse images Luma releases Ray3 video model with self-evaluation claims and 16-bit HDR EXR output Caira camera app adds Gemini Omni and Runway Aleph video editing in private beta Google launches agentic video understanding for Gemini Flash, claiming up to 88% fewer video tokens World Labs unveils Atlas, a 1440p video model with pixel-level camera control Sand.ai open-sources 114B-parameter MAGI Preview video model with 6B active MoE parameters World Labs unveils Atlas, a spatial world model that generates 1440p videos up to one minute xAI's Grok Imagine Raises Reference Limit From 7 to 14 for Video Generation Runway previews GWM Worlds 2 for real-time interactive 720p video and audio worlds fal releases H3 Max, a post-trained MiniMax H3 video model claiming #1 benchmark rankings Google adds agent-based video retrieval to Gemini Flash, claiming 88% lower token use fal releases H3 Max, a post-trained MiniMax H3 video model claiming 3-second generation for 5-second clips World Labs releases Atlas for controllable 1440p, one-minute 3D world generation fal Releases H3 Max, a Post-Trained MiniMax H3 Video Model Claiming #1 Benchmark Ranks Google Adds Agentic Video Understanding to Gemini Flash Models fal Releases H3 Max, a Post-Trained MiniMax H3 Video Model That Generates Five Seconds in About Three World Labs releases Atlas world model with 3D reconstruction and camera trajectory control fal releases H3 Max, a post-trained MiniMax H3 video model that generates 5-second clips in about 3 seconds Google adds agentic video analysis to Gemini Flash models, claiming 88% lower token use MiniMax H3’s public vLLM serving stack renders 10.125 seconds of audiovisual video in under nine seconds World Labs launches Atlas, a shared model for controlled video, 3D reconstruction and robot views Visko launches Orbis for continuous real-time video generation with live prompt changes Runway Solaris Generates Stateful App Interfaces Frame by Frame Without Code World Labs unveils Atlas, a multimodal world model that generates controllable 1440p videos up to one minute Google adds agentic video understanding to Gemini Flash models, claiming up to 88% lower token use Vidu Announces S1 Voice-Driven Real-Time Interactive Video Model Visko opens Orbis, a real-time 4K interactive video model, after $10M pre-seed round Visko opens Orbis, a live AI video model that streams interactive 4K worlds for hours fal releases H3 Max, claiming top video benchmark rankings and 3-second generation for 5-second clips FastH3 open-weight preview cuts MiniMax H3 video and audio generation to four DiT calls Gemini Omni 1.1 Flash Adds 40-Second Scene Extension and First-to-Last-Frame Video Control Alibaba Wan 3 unifies four video modes, adds 30-second generation and native audio Overworld releases Waypoint-1, a real-time video diffusion model controlled by text, mouse, and keyboard NVIDIA releases Cosmos 3 open multimodal world model for physical AI Alibaba launches Wan3.0 AI video model for 30-second document-based clips Runway releases Gen-4.5, claiming top rank on Artificial Analysis' text-to-video benchmark Google releases Veo 3.1 Lite in paid preview at under half the cost of Veo 3.1 Fast FastVideo releases open-weight 4-step FastH3 Preview v1 for MiniMax H3 video and audio generation Reported Gemini Omni 1.1 Flash update chains 10-second clips into 40-second scenes with 4K output Google's Gemini Omni 1.1 Flash adds 40-second video extension, frame pinning, and 4K output Google releases Gemini Omni 1.1 Flash with 40-second scene extension and 4K upscaling Google Gemini Omni 1.1 Flash adds 4K output, 40-second scene extension, and keyframe controls Google releases Gemini Omni 1.1 Flash with video continuation, frame controls, and 4K upscaling Google makes Gemini Omni 1.1 Flash generally available with scene extension and 4K output Alibaba releases Wan3.0, a 30-second video model that takes documents and web pages as inputs Z.ai open-sources 320B-parameter GLM-5.3-Flash for million-token video inputs Google Omni 1.1 Flash adds 1080p and upscaled 4K video output Google's Gemini Omni 1.1 Flash Reportedly Adds Direct 4K Video Output and 40-Second Extensions Gemini Omni 1.1 Flash adds 4K upscaling, 40-second scene extension, and frame-to-frame video generation FastVideo releases FastMetal-QAD video models for local Apple Silicon generation Google releases Gemini Omni 1.1 Flash with 40-second scene extension and 4K video upscaling Alibaba opens Wan3.0 beta with 30-second video generation and document-to-video input Google releases Gemini Omni 1.1 Flash with 40-second scene extension and 4K upscaling JD.com Open Sources Echo-WM, a Navigable Video, Audio, Music and Speech World Model Alibaba WAN 3.0 raises single-pass video length to 30 seconds, adds document and web inputs Alibaba releases Wan3.0 video model after $10 billion share placement Alibaba Releases Closed-Weight Wan3.0 Video Model With 30-Second 1080p Generation Alibaba's Wan3.0 adds 30-second video, audio, and document inputs at $6 per 1080p clip 4DAnyone generates 16 consistent views from one video for 4D human reconstruction Alibaba's Wan 3.0 beta claims 30-second single-pass AI video and document inputs Alibaba rolls out Wan 3.0 with 30-second 1080p document-to-video generation Alibaba releases Wan3.0 with 30-second generation and document-to-video input Alibaba Rolls Out Wan3.0 Video Model With 30-Second Generation and Document Inputs Alibaba launches Wan3.0 video model with 30-second generation and document inputs

5 alerts

My alerts

Sign in to create alerts.

All alerts

AI video model releasesby User 6432 Duneby 제욱 AIby 제욱 App Store 5.2.3by 주현 해커뉴스by 성현