Chinese labs claim nine of the top 10 text-to-video spots

The chatbot wars get the headlines, but the clearest evidence of China's AI lead is showing up on a different leaderboard: video generation. Nine of the world's ten best text-to-video models are now Chinese, and the analysts watching say the lead points somewhere bigger than Hollywood.

Chinese AI labs hold nine of the ten top-ranked text-to-video models on Artificial Analysis' leaderboard, with Google's Gemini the only non-Chinese system in the top tier, Bloomberg reports. Writing over the weekend, Bloomberg's Catherine Thorbecke said the models are gaining global adoption and argued the video lead could give Chinese labs an edge in building world models — the AI systems that learn to simulate environments, widely seen as a route toward more general intelligence. Artificial Analysis' blind arena standings back the framing: Google's Gemini Omni Flash leads the text-to-video bracket, with MiniMax's H3 just behind, and ByteDance's Seedance 2.0, Alibaba's Wan and Kuaishou's Kling 3.0 filling out the top ranks.

The leaderboard is the endgame of a year-long video war the US largely ceded. OpenAI's Sora — the model that ignited the text-to-video moment in 2024 — was shelved after the company balked at the operating costs of a resource-intensive service, according to the LA Times. Chinese labs stepped into the void with models that ship faster and iterate harder, and the capital followed: Kling recently pulled a $2.8 billion financing round backed by Alibaba and Tencent, while Alibaba has kept shipping its own Wan video line in parallel. The result is a rare category where Chinese labs don't just match US output — they set the benchmark everyone else is measured against.

Why it matters beyond entertainment: video generation is increasingly treated as a training ground for world models, and whoever owns the best video architectures may own the foundation for the next phase of AI. If the world-model era gets built on Chinese video stacks, the industry's center of gravity shifts another step away from Silicon Valley — the same pattern already playing out across chips, open-weight models and price. For US labs, the video category is now the sharpest demonstration of the gap, and the open question is whether anyone builds a serious challenger while Sora sits on ice.

What to watch: whether OpenAI revives a video push before the world-model race is decided, and whether Google's Gemini can hold the top spot against the Chinese cluster.

If Chinese labs own the video stack, do they also own the path to world models — or is the race still wide open? Tell us in the comments.

Kling 4.0 goes live with 30-second native video at 10-bit HDR

Kuaishou's Kling AI has spent two years becoming the video model people actually finish projects with. Its new flagship, Kling 4.0, went fully live on October 15 — and this round is less about prettier frames than about the two things that make a generated shot usable in a real edit: length and control. Kling 4.0 generates up to 30 seconds in a single native pass, outputs 10-bit HDR at 4K, 1080p and 21:9, and takes up to 15 reference assets in one task. That budget splits across ten images, fi

Brain organoids classified Braille at 83% — in a trio, not alone

Two stories today: human brain tissue on a chip decoded a tactile signal, and the data-centre company behind some of the largest AI contracts in Europe turns out to be six months behind on its flagship site. Human forebrain organoids grown on a microelectrode array classified Braille letters with 61 percent accuracy on their own and 83 percent when the responses of three organoids were combined, in a peer-reviewed study from the University of Bristol. The team took event-based tactile readings

Suleyman says Microsoft's base model beat OpenAI's on a quarter compute

Microsoft finished pre-training its biggest in-house model yet — then used it to make a point about the company it bankrolls. Microsoft AI completed pre-training of MAI-Base-2, a mixture-of-experts foundation model with 80 billion active and 2.8 trillion total parameters, and says the recipe is 4.2–12.3x more pre-training FLOPs-efficient than DeepSeek-V4-Pro-Base. The run itself is the number worth keeping: 16,000 Nvidia GB200 GPUs for 38 days across August and September, at 21% average model

Amazon's new Bee devices listen all day — for $11.99 a month

Amazon turned its always-listening Bee line into a two-device family on Thursday, and NVIDIA opened a 2.8-trillion-token Indic corpus recipe. The connecting thread: both are bets that the next AI money is in memory — yours, or your model's. Amazon put two new always-listening devices on sale, and both now lead to a subscription. Bee Frontier, the next-generation wristband that also clips to clothing, ships now at $74.99 in Onyx, Lavender and Sage — up from the $50 predecessor — with seven-day

Read more

Kling 4.0 goes live with 30-second native video at 10-bit HDR

Brain organoids classified Braille at 83% — in a trio, not alone

Suleyman says Microsoft's base model beat OpenAI's on a quarter compute

Amazon's new Bee devices listen all day — for $11.99 a month