Same model, different harness, different result. Octomind, an open-source Rust coding agent, ran GLM-5.2 on 25 tasks rebuilt from real merged PRs and solved 24/25. OpenCode with the same model and endpoint solved 19/25, at about 2x the cost.
Muvon/octomind
15 views
Octomind – open-source AI harness that saves your tokens, static rust
Octomind – open-source AI harness that saves your tokens, static rust Muvon/octomind (news.ycombinator.com)
1 like · 20 views
Octomind – open-source AI harness that saves your tokens, static rust Muvon/octomind (news.ycombinator.com)
I'm not sure if I missed something, but the benchmarks [0] look outdated since they use Claude Opus 5 (instead of Opus 5.5), GPT 5.6 Sol (instead of 6.1 Sol), and GLM 5.2 (instead of DeepSeek V4.1 Flash 1017, which is the most popular model on OpenRouter currently [1]). I also cannot find any information regarding thinking efforts for each.
Since these benchmarks were done in July, they were probably using older versions of Claude Code, Codex, and OpenCode as well.
Would appreciate if the benchmarks are updated, and the thinking efforts and the versions of Claude Code, Codex, and OpenCode are specified.
[0] Muvon/octomind
[1] openrouter.ai
Octomind – open-source AI harness that saves your tokens, static rust Muvon/octomind (news.ycombinator.com)
1 like · 17 views
Caveat: it's their own benchmark (octobench, Jul 30). But per-case results + a repro guide are public.
Same test: Claude Code + Opus 5 solved 23/25 for $82. Codex + gpt-5.6-sol solved 21/25 for $15, fastest.
Repo: Muvon/octomind
1 like · 44 views
Octomind – open-source AI harness that saves your tokens, static rust
#HackerNews
Muvon/octomind
Octomind – open-source AI harness that saves your tokens, static rust Muvon/octomind (news.ycombinator.com)
Octomind – open-source AI harness that saves your tokens, static rust
Muvon/octomind
Mentions
Same model, different harness, different result. Octomind, an open-source Rust coding agent, ran GLM-5.2 on 25 tasks rebuilt from real merged PRs and solved 24/25. OpenCode with the same model and endpoint solved 19/25, at about 2x the cost. Muvon/octomind
Octomind – open-source AI harness that saves your tokens, static rust
Octomind – open-source AI harness that saves your tokens, static rust Muvon/octomind (news.ycombinator.com)
Octomind – open-source AI harness that saves your tokens, static rust Muvon/octomind (news.ycombinator.com)
I'm not sure if I missed something, but the benchmarks [0] look outdated since they use Claude Opus 5 (instead of Opus 5.5), GPT 5.6 Sol (instead of 6.1 Sol), and GLM 5.2 (instead of DeepSeek V4.1 Flash 1017, which is the most popular model on OpenRouter currently [1]). I also cannot find any information regarding thinking efforts for each. Since these benchmarks were done in July, they were probably using older versions of Claude Code, Codex, and OpenCode as well. Would appreciate if the benchmarks are updated, and the thinking efforts and the versions of Claude Code, Codex, and OpenCode are specified. [0] Muvon/octomind [1] openrouter.ai
Octomind – open-source AI harness that saves your tokens, static rust Muvon/octomind (news.ycombinator.com)
Caveat: it's their own benchmark (octobench, Jul 30). But per-case results + a repro guide are public. Same test: Claude Code + Opus 5 solved 23/25 for $82. Codex + gpt-5.6-sol solved 21/25 for $15, fastest. Repo: Muvon/octomind
Octomind – open-source AI harness that saves your tokens, static rust #HackerNews Muvon/octomind
Octomind – open-source AI harness that saves your tokens, static rust Muvon/octomind (news.ycombinator.com)
Octomind – open-source AI harness that saves your tokens, static rust Muvon/octomind