Laguna S 2.1（118B 参数）在游戏编程任务中达到与 753B 的 GLM-5.2 同等的质量，参数量仅为后者的 1/6。在构建三款自玩街机游戏时，Laguna S 2.1 仅用 10.3K tokens 完成，而 GLM-5.2 需要 26.4K tokens。Laguna S 2.1 是三者中唯一能在 128GB MacBook 上本地运行的模型。
Laguna S 2.1 (118B param) looks so great for intelligence per token.
It matched the 753B GLM-5.2 on game coding with 6x fewer parameters.
Both models faced the same job, alongside a third contender named Hy3.
Test was done on atomic[.]chat, a desktop app that runs LLMs locally.
Each had to build 3 arcade games as self-contained HTML files.
The targets were Geometry Dash, Doodle Jump, and Air Hockey. Every game also had to play itself through a built-in bot. So no human touches the controls once the file opens.
The smaller Laguna S 2.1 produced its solutions using just 10.3K tokens total. GLM-5.2 needed 26.4K tokens to reach roughly the same result.
Laguna S 2.1 performs at GLM-5.2 level on building popular games with 6x fewer params! We gave three local models the same task: build three popular arcade games that play themselves. Each game is one self-contained HTML file with a bot that plays it. Prompts: – Geometry Dash – Doodle Jump – Air Hockey Outputs: Laguna S 2.1: 10.3K tokens GLM-5.2: 26.4K tokens Hy3: 10.4K tokens Laguna held its quality against a 753B model. We think Laguna's Geometry Dash looked the best of the three, the cube clears every spike and block. GLM won Air Hockey. Its table looked the most detailed of all. Hy3 was the only model that added shooting to its Doodle Jump. But Laguna is the only model in our benchmark that runs on a MacBook with 128GB!在 X 查看被引用的帖子
来源：Rohan Paul · x.com