Chubby♨️@kimmonismus
57AI 编辑部评分,满分 100
2026-08-04 00:43· 39分钟前
跳到正文
AI 摘要

阿里Qwen3.8-Max据称自主运行16天,产出265个提交、127个PR和151个issue,输出成本较GPT-5.6 Sol低80%、较Claude Fable 5低88%。该2.4T参数MoE模型输入/输出价格分别为$2/$6每百万token,PaperBench评分领先上述两者,开源权重下周发布。

The leaps in capabilities are literally insane.

Open Source Opus 4.8, according to DeepSWE Benchmark.

The jump to Qwen 3.7 is simply insane.

Chubby♨️Holy, China strikes again: Qwen3.8-Max reportedly worked autonomously for 16 days while costing 80% less than GPT-5.6 Sol and 88% less than Claude Fable 5 on ou...
Chubby♨️ · @kimmonismus · X·2026-08-04 00:43·39分钟前
在 X 看原推· x.com
AI 摘要

阿里Qwen3.8-Max据称自主运行16天,产出265个提交、127个PR和151个issue,输出成本较GPT-5.6 Sol低80%、较Claude Fable 5低88%。该2.4T参数MoE模型输入/输出价格分别为$2/$6每百万token,PaperBench评分领先上述两者,开源权重下周发布。

The leaps in capabilities are literally insane.

Open Source Opus 4.8, according to DeepSWE Benchmark.

The jump to Qwen 3.7 is simply insane.

Chubby♨️Holy, China strikes again: Qwen3.8-Max reportedly worked autonomously for 16 days while costing 80% less than GPT-5.6 Sol and 88% less than Claude Fable 5 on ou...
在 X 查看原推x.com