Skip to content
IFAI

Alibaba

Company61 stories

  1. ResearchReach 59

    GPT-6 Astra rewrites neural network weights directly without training

    Researchers introduced Weight-Space Programming, showing OpenAI's GPT-6 Astra can natively interpret and tweak neural network weights without traditional gradient training. By altering just 0.016% of Qwen3.5-0.8B's weights, it fixed a reasoning loop and bumped math accuracy from 7.9% to 48.1%. Why it matters: Direct parameter steering hints at automated model patching, though current results rely heavily on external Python tools rather than pure weight intuition.

  2. Generative mediaReach 41

    Alibaba launches Qwen-Image-2.1 APIs for multi-tool visual workflows

    Alibaba just rolled out commercial Pro and Turbo APIs for Qwen-Image-2.1 on its Model Studio platform. The single 7B model collapses separate design steps by handling generation, 10-image reference edits, native RGBA transparency, and 2K typography. Turbo starts at around $0.014 per image, giving production teams a full pipeline without self-hosting GPUs.

  3. Generative mediaReach 45

    Alibaba's new open image model generates 2K visuals in just 8 steps

    Alibaba just open-sourced Qwen-Image-2.1-Turbo, an accelerated 7B image generation and editing checkpoint that produces crisp 2K visuals in only 8 denoising steps. It handles both prompt-based generation and natural-language edits without trading away visual quality. The weights are up on Hugging Face and ModelScope alongside official hosted APIs, making high-res local image generation substantially faster.

  4. AI toolsReach 40

    Alibaba drops agent-powered computer and smart wearables

    Alibaba just unveiled an agentic hardware lineup at the 2026 Apsara Conference, headlined by Qwen Book, a computer running an agent-native OS that lets models control apps and cloud tools. The launch also includes eye-tracking Qwen Glasses, Bose-partnered AI earbuds, and the QwenNote A2 voice recorder. It marks a major push to embed autonomous AI agents straight into personal devices.

  6. AI modelsReach 53

    Alibaba drops Qwen3.8-Flash: 125B open-weight MoE with 6B active params

    Alibaba just launched Qwen3.8-Flash, an open-weight multimodal MoE model with a 125B backbone that activates just 6B parameters per token. It handles 262K native context, previews the upcoming Qwen4 architecture, and cuts API input costs to $0.16 per million tokens while matching heavyweights in coding benchmarks. Weights are live on Hugging Face and ModelScope.

  7. Open sourceReach 48

    BreezeBlue drops Breeze-TTS-2 with 40ms real-time streaming

    BreezeBlue just released open weights and inference code for Breeze-TTS-2, a bilingual text-to-speech model built for real-time interaction. It delivers sub-40ms time-to-first-audio on an H100 and lets developers design distinct voices from scratch using natural language prompts or inject vocal cues like sighs and laughs directly into transcripts. The code is Apache 2.0, while model weights remain research-only.

  8. Generative mediaReach 45

    Alibaba drops HappyShrimp 1.0 to turn single text prompts into full songs

    Alibaba just rolled out HappyShrimp 1.0 into beta, an AI music generation model that crafts complete tracks—melody, arrangement, lyrics, and vocals—from a single text description. It maps out musical structure and emotional tone across genres like pop, rock, and Chinese style, with China's Taihe Music already on board for artist co-creation.

  9. Open sourceReach 56

    Alibaba's Qwen models top 3B downloads, outpacing Meta and Google

    Alibaba's Qwen models have just crossed 3 billion global downloads over the past six months, overtaking Meta and Google as the world's leading open-weight AI family according to Hugging Face. In 2026, Google saw 418 million downloads and Meta reached 227 million. With over 460 open models and 300,000 community derivatives, Alibaba is rapidly cementing itself as the backbone of open-source AI.

  10. Open sourceReach 51

    Alibaba just open-sourced Qwen3.8-27B under Apache 2.0

    Alibaba just released the open weights for Qwen3.8-27B, a native multimodal dense model under the Apache 2.0 license. At 27B parameters, it outperforms Qwen3.7-Plus and features a 262K native context window extendable to 1M tokens. Open weights for the Max-level Qwen3.8-2.4T-A95B are also available on Hugging Face and ModelScope.

  11. Generative mediaReach 42

    Alibaba's Wan3.0 turns PPTs into 30-second AI videos

    Alibaba Cloud just launched Wan3.0, extending single-pass video generation to 30 seconds. For the first time, the model can ingest office files like PDFs, slides, and spreadsheets alongside text and images to produce narrative videos. It also sharpens facial fidelity and adds built-in video editing, with API pricing starting at $0.05 per second.

  12. AI toolsReach 39

    Alibaba launches QwenWork, an all-in-one workplace AI agent suite

    Alibaba just kicked off the public beta of QwenWork in China, an all-in-one workplace AI agent platform tying together desktop, cloud, and DingTalk. It packs autonomous agents, native multimodal generation, and instant web app creation, with access to the flagship Qwen3.8 Max model starting August 3. A standalone mobile app and international rollout are also on the roadmap.

  13. AI modelsReach 52

    Alibaba drops Qwen3.8-Max with 2.4T parameters

    Alibaba just launched Qwen3.8-Max, its most powerful flagship multimodal model featuring 2.4 trillion total parameters while activating just 95 billion via sparse MoE. It supports a 1M token context window and ranks second in Vision Arena. The model is accessible via API on Alibaba Cloud Model Studio, with weights releasing next week.

  15. Open sourceReach 48

    Z-Anime brings full fine-tuning to open-source anime art

    Developer SeeSee21 just dropped Z-Anime, a 6B parameter diffusion model built by fully fine-tuning Alibaba's Z-Image Base architecture instead of stacking LoRAs. It features an FP8 build under 6GB and a distilled 4-step version that churns out images in seconds on an 8GB GPU. Unlike typical anime models, it demands natural language prompts instead of comma-separated tags.

  16. Generative mediaReach 39

    Alibaba's Qwen-Image-2.0-Pro cracks the top 10 on Arena

    Alibaba just launched Qwen-Image-2.0-Pro, its latest text-to-image model featuring upgraded image quality, multilingual text rendering, and sharper instruction following. The new model debuted at #9 worldwide on Arena's Text-to-Image leaderboard, also cracking the top 10 for portraits and photorealistic styles. It is available to test on ModelScope alongside an official API.

  17. Open sourceReach 66

    Alibaba drops Qwen3.6-27B to punch above its weight in coding

    Alibaba just open-sourced Qwen3.6-27B under the Apache 2.0 license, packing flagship-grade agentic coding into a 27B dense model. According to the team, it outperforms their much larger Qwen3.5-397B-A17B across major coding benchmarks while supporting both thinking and standard inference modes. It proves small dense models are catching up fast to monster MoE architectures in specialized developer workflows.

  18. AI modelsReach 49

    Alibaba teases Qwen3.6-Max with big agentic coding boost

    Alibaba just released an early preview of Qwen3.6-Max-Preview, offering a first look at its next flagship model. Compared to Qwen3.6-Plus, it brings major leaps in agentic coding, instruction following, and real-world agent reliability. It is Alibaba's latest push to make autonomous agents dependable for complex tasks.

  19. Open sourceReach 57

    Alibaba open-sources Qwen3.6-35B-A3B sparse MoE model

    Alibaba just open-sourced Qwen3.6-35B-A3B under Apache 2.0, a sparse MoE model packing 35B total parameters with only 3B active. Despite its tiny active footprint, it claims agentic coding performance on par with models 10x its active size, alongside multimodal reasoning and thinking modes.