Nathan Lambert@natolambertQwen 3.8 Max 的表现值得关注，因为此前他们的大模型在 API 端采用率并不高，现在（在许可条款允许的前提下）可能会有更多人关注。这或许正是 Qwen 团队追赶顶级模型所需的反馈循环。 定价、许可、持续发布、补丁反馈等，即将成为 Kimi、GLM、Qwen、DeepSeek 等厂商之间的分水岭——它们都有能力推出领先模型，也都在不同细分市场各有斩获。
It'll be interesting to see how Qwen 3.8 Max lands， as their previous large models didn't get a ton of adoption in the API， more people will likely look now （pending license）. This could be the feedback loop the Qwen team needed to catch up on the biggest models.
Pricing， license， consistent releases， patching feedback， etc about to be the boundary between Kimi， GLM， Qwen， DeepSeek， etc. they're all capable of landing the leading model and all have been winning different pieces of the market.
Qwen 3.8 Max 的表现值得关注，因为此前他们的大模型在 API 端采用率并不高，现在（在许可条款允许的前提下）可能会有更多人关注。这或许正是 Qwen 团队追赶顶级模型所需的反馈循环。 定价、许可、持续发布、补丁反馈等，即将成为 Kimi、GLM、Qwen、DeepSeek 等厂商之间的分水岭——它们都有能力推出领先模型，也都在不同细分市场各有斩获。
It'll be interesting to see how Qwen 3.8 Max lands， as their previous large models didn't get a ton of adoption in the API， more people will likely look now （pending license）. This could be the feedback loop the Qwen team needed to catch up on the biggest models.
Pricing， license， consistent releases， patching feedback， etc about to be the boundary between Kimi， GLM， Qwen， DeepSeek， etc. they're all capable of landing the leading model and all have been winning different pieces of the market.