NAll news

All news

Sunday12 items

  5. Tongyi Qianwen / QwenFirst-partyModel releases
    AI score73

    Qwen Open-Sources Qwen-Image-2.1 Image Generation and Editing Model

    Qwen has released Qwen-Image-2.1 and opened its weights. This is an image generation model in the Qwen-Image series that balances performance and cost, with unified support for both generation and editing. The model uses a 7B lightweight architecture, can natively generate and edit RGBA transparent layers, supports up to 10 reference images and precise local editing, and covers scenarios such as panoramas, infographics, and virtual try-on.

    Same story · 4 reports

    #Model releases#Image generation#Open-source ecosystem

Saturday3 items

  1. Tongyi Qianwen / QwenFirst-partyModel releases
    AI score62

    Qwen Releases Real-Time Simultaneous Interpretation Model Qwen3.8-LiveTranslate

    Qwen has released its next-generation real-time simultaneous interpretation model, Qwen3.8-LiveTranslate, built on the Interleave architecture. Across 60 languages, it reduces average latency (LAAL) from 2.8s to 2.3s, while improving faithfulness, fluency, and conciseness. New capabilities include real-time speaker diarization, which can distinguish speakers in multi-party speech and preserve each person's voice through more stable voice cloning; synchronized bilingual display, presenting the source text and translation on the same screen; and long-context disambiguation, which uses conversation history to clarify names and terminology to ensure translation consistency. The model is now available on QwenCloud, and the official release also provides links to a blog post and case studies.

    #Model releases#Speech#Multimodal

Friday5 items

  3. Tongyi Qianwen / QwenFirst-partyModel releases
    AI score80

    Alibaba Qwen Releases Omni-Modal Model Qwen3.8-Omni-Flash

    Alibaba Qwen has released Qwen3.8-Omni-Flash, its first omni-modal model built around agent capabilities, with native support for audio-visual understanding, reasoning, and tool calling. The company says its audio-visual capabilities are close to Gemini 3.8 Flash, with agent performance improving by an average of 19.5 points on WildClawBench-MM and UniClawBench; it supports a 1M token context and uses 51.8% fewer tokens than static understanding on OmniVideoBench. Video input costs are about 89% lower than Qwen3.5-Omni-Plus, while Qwen-MM-Plugins is being open-sourced and Qwen-Live Harness is coming soon.

    #Model releases#Multimodal#Agents

Thursday2 items

  1. StepFunFirst-partyModel releases
    AI score60

    StepFun Releases StepAudio 3 Music Music Generation Foundation Model

    StepFun and ACE Studio jointly release StepAudio 3 Music, StepFun's first music generation foundation model, which can generate complete songs from prompts and lyrics. The model integrates four workflows: song generation, accompaniment generation, music cover, and vocal-to-song arrangement. ABC-COT plans the musical structure and arrangement before synthesis, and users can rewrite prompts, edit the ABC notation, and iterate. A Hugging Face Space demo and API documentation are now available, and the ABC-COT API is coming soon.

    #Model releases#Speech#Hugging Face