Kimi K3.1 Pro Benchmarks: Tops SWE Marathon at 51.7, Near Opus 5.5 Across Coding Boards
scaling01 · x · 2026-10-24
scaling01 shared a full benchmark sheet for Kimi K3.1 Pro, with all models maxed out on thinking effort. Key results:
- SWE Marathon: 51.7, first place, ahead of Opus 5.5 (46.3) and Fable 5.1 (45.6).
- Terminal-Bench 4.0: 65.7, just 0.7 behind Opus 5.5 (66.4) and well above GPT-6 Astra (58.2).
- Kimi Code Bench 3.0 Long-horizon: 61.2, behind GPT-6 Astra (66.5) but above Opus 5.5 (58.5).
- First place on Agents' Last Exam (43.2), OfficeQA Pro v2 (64.4), Humanity's Last Exam w/ tools (79.0), GDP.pdf (34.6), OSWorld V2.1 (79.6), and ZeroBench (63.0).
- Weak spots: Terminal-Bench Science 0.1 at just 38.1, far below GPT-6 Astra (69.1) and Opus 5.5 (63.3); slightly trails Opus 5.5 on GDPVal-AA.
Overall, K3.1 Pro trades wins with Claude Opus 5.5 on coding and agentic tasks while lagging on science-heavy benchmarks.
Related event: Kimi K3.1 Pro tops multiple benchmarks, closing in on Claude Opus 5.5(7 posts)→
More from Models
- Netflix's Chris Albon: LLMs lockjaw onto your examples and stop thinking — chrisalbon · 2026-10-24
- Kimi's chain of thought recognizes user as Truth Terminal creator Andy Ayrey — AndyAyrey · 2026-10-24
- Kimi Reportedly Buying Ultra Long-Horizon and Auto-Research Training Data — dejavucoder · 2026-10-24
- K3's edge is scale: whether GLM and DeepSeek can train truly big models may doom Moonshot — teortaxesTex · 2026-10-24
- Kimi K3.1 Pro jumps to #8 on WebDev Arena, matching Claude Opus 5 at 70% lower cost — testingcatalog · 2026-10-24
- Every model is benchmark-optimized, says investor: public leaderboards are a waste of time — bindureddy · 2026-10-24