Alibaba Updates Qwen3.8 Max Preview and Confirms Open Weight Release of 2.4 Trillion Parameter AI Model
Alibaba updated its Qwen3.8-Max-Preview generative AI model on Sunday, noting broad performance gains and improved web frontend capabilities ahead of the official launch. The company announced that the final version of the 2.4T parameters model will be released with open weights for public access.
This marks a departure from the closed-weight strategy used for prior Qwen Max models, as the firm targets developer adoption through OpenAI- and Anthropic-compatible APIs and a trial period charging a fraction of standard credits. The model processes text, images, video, and documents, though it currently lacks published benchmark data despite performance claims positioning it second to Anthropic’s Fable 5. Alibaba’s Hong Kong shares rose more than 5% on the news.
From the sources (25 posts)
@alibaba_qwenQwen3.8 is launching and going open-weight soon!🌐 With a massive 2.4T parameters, this model is continuously evolving. We believe it’s one of the most powerful model available today, compatible to leading frontier AI models , second only t
@alibabagroupMeet Qwen3.8-Max-Preview, our latest LLM with 2.4 trillion parameters, fully multimodal capabilities, and sharper performance in code engineering and professional cowork. Now live on Alibaba's Token Plan, Qoder and QoderWork, explore it to
@synthwaveddQwen are testing Qwen-3.8-Max-Preview, now available for Qwen Token Plan subscribers in mainland China. The model is the largest ever trained by Qwen at 2.4T total parameters. Kimi, Qwen, DeepSeek, GLM... China are cooking!
@synthwaveddQwen are claiming Qwen 3.8 Max is "second only to Fable." That would make it better than K3, only days after K3 launched. Wow.
@zephyr_z9Another big boi model
@zephyr_z9"Qwen3.8 is launching and going open-weight soon" Interesting, so the big 2.4T model from Qwen will be open source
@bdsqlszRT @Alibaba_Qwen: Qwen3.8 is launching and going open-weight soon!🌐 With a massive 2.4T parameters, this model is continuously evolving. W…
@elliotarledgeRT @Alibaba_Qwen: Qwen3.8 is launching and going open-weight soon!🌐 With a massive 2.4T parameters, this model is continuously evolving. W…
@theahmadosmanLooks like we're getting Qwen 3.8 weights folks!!!
@kimmonismusHoly, Qwen 3.8 supposedly ahead of GPT-5.6 and only slightly behind Fable 5! - 2.4t Parameters - Open Source / Open Weight - full release soon, already available for testing as Qwen 3.8 max-Max-Preview What the frick, such insane release
@teortaxestex2.4T is a design Chinese hyperscalers like. Ernie 5 and ByteDance's experiments also used this number. Now Alibaba.
@mtsliveSITUATION DETECTED: Alibaba has released Qwen3.8, a new 2.4T parameter model that they claim is behind only Claude Fable 5 in performance.
@victormustarRT @Alibaba_Qwen: Qwen3.8 is launching and going open-weight soon!🌐 With a massive 2.4T parameters, this model is continuously evolving. W…
@dejavucoder
@suchenzanghow dare they give us the model for free after stealing from america! the audacity! the horror! such foolish decels! such blindness to agi! such short-sightedness to not withhold these dangerous capabilities from the evil plebs and nefa
@thezachmuellerThe age of single B200 node models is officially over. Was ballparking November myself. Oops
@algo_diverRT @Alibaba_Qwen: Qwen3.8 is launching and going open-weight soon!🌐 With a massive 2.4T parameters, this model is continuously evolving. W…
@mark_kYou thought Kimi K3 was the last surprise for a while? WRONG. Qwen 3.8 is launching soon, and it will be open-weight. With a massive 2.4T parameters, Alibaba claims it is one of the most powerful models available today, competitive with the
@ben_burtenshawRT @Alibaba_Qwen: Qwen3.8 is launching and going open-weight soon!🌐 With a massive 2.4T parameters, this model is continuously evolving. W…
@hosseebWow, Alibaba coming in hot. Dropped on a Sunday, no benchmarks, no weights yet, just a claim that "this is second only to Fable, come try it via our API." After Qwen team underwent a massive restructuring in March, Qwen 3.7, their previ
@synthwaveddBeen testing Qwen 3.8 Max Preview. While it's a decent step up over 3.7, it simply doesn't feel as good in real-world use as K3 does. Qwen consistently underperform in real-world use versus benchmark scores, and it seems that remains the c
@teortaxestex> We're looking forward to a more capable, official version — and to open-weight it for everyone.🚀 this is actually more than I expected. They are announcing open-weighting of the final 3.8 Max, not this preview.
@synthwaveddRT @Alibaba_Qwen: During Preview, Qwen3.8 is getting better by the day. Latest version is live now, with broad gains and a big step up on w…
@alibaba_qwenDuring Preview, Qwen3.8 is getting better by the day. Latest version is live now, with broad gains and a big step up on web frontend. Thank you all — the response to Qwen3.8-Max-Preview blew us away. 🫶🫶 Qwen3.8 is still evolving daily. Co
@poezhao0605Alibaba's Qwen team(@Alibaba_Qwen) is making its boldest claim yet: Qwen3.8-Max is second only to Anthropic's Fable 5. The full benchmarks are still to come. The distribution strategy is already clear. Qwen3.8-Max is a 2.4T-parameter mult