Alibaba's Qwen team is previewing the Qwen4 architecture with Qwen3.8-Flash-Next, a mixture-of-experts model that activates just 6 out of 125 billion parameters per token. At one-ninth the training cost, it beats much larger competitors like DeepSeek-V4-Flash and Claude Opus 4.6 on coding and office benchmarks, adding more pricing pressure on OpenAI and Anthropic.

The article Alibaba releases Qwen3.8-Flash-Next, targeting "ultimate cost efficiency" appeared first on The Decoder.

XPeng – previously known only for electric cars – has raised more than $900 million for its robotics business at a post-money valuation of more than $6.3 billion, as the Chinese electric vehicle maker prepares to begin mass production of its IRON humanoid robot. The company says the financing is the largest single-round private capital […]