为什么瞧不起 Luna 反而 IDA 官方默默地夸奖 Luna

Decidable6471 · 2026-09-28T01:22:00+08:00

hex-rays.comHex-Rays IDA MCP ServerMeet the official IDA MCP server: free, open source, and model-agnostic. Connect IDA to AI agents and cut token use by ~20% with Code Mode.重点不是 MCP 的问题里面有段话```According to our internal, malware analysis-focused benchmarks, we’ve found that DeepSeek v4.1 Flash, Kimi K3, and GLM 5.3 perform on a level similar to GPT-6-Luna and Gemini 3.8 Flash. Unfortunately, they do require dedicated hardware. For “prosumer grade” hardware (128-256GB VRAM), DeepSeek V4 Flash 0731 and GLM 5.3 Flash also give decent results.```

Decidable64712026-09-28
11830
hex-rays.com
Hex-Rays IDA MCP Server
Meet the official IDA MCP server: free, open source, and model-agnostic. Connect IDA to AI agents and cut token use by ~20% with Code Mode.
重点不是 MCP 的问题
里面有段话```
According to our internal, malware analysis-focused benchmarks, we’ve found that DeepSeek v4.1 Flash, Kimi K3, and GLM 5.3 perform on a level similar to GPT-6-Luna and Gemini 3.8 Flash. Unfortunately, they do require dedicated hardware. For “prosumer grade” hardware (128-256GB VRAM), DeepSeek V4 Flash 0731 and GLM 5.3 Flash also give decent results.
```
主楼
None2026-09-28
15475
哈? 这说明了什么?
只说明了这活 luna 也能干, 它的小标题还是 IDA MCP 与模型无关
#1
木原金2026-09-28
12283
翻译：IDA的内部基准测试还未公布就已经饱和
#2
Belope2026-09-28
23767
luna 对我来说无论 5.6 还是 6 都太慢了，比我公益站基本无限的 dsv4f /glm5.3f 还慢这个程度，其实隐性成本在我这里已经不比 6sol 低了。
#3

发表回复

登录后回复