RLTasksby ulam.ai
Models / Kimi

Kimi / MODEL EVIDENCE

Kimi K2.5

77 benchmarks · 114 observations · 53 sources · released 2026-01-27

Scores at a glance

77 benchmarks · large dot: Kimi K2.5; small dots: closest models where reported

V* ↗86.9
86.498.4

Kimi K2.5 (thinking) · V* (w/ python) ↗

MMMU ↗84.3
80.687.2

K2.5-1T-A32B · MMMU · Qwen report ↗

Vals RSI Index ↗0.00 reference score
00.46

Kimi K2.5 (max) · Compression · reference-anchored score ↗

3 more reported results

0.25 reference score · Kimi K2.5 (max) · LM training · reference-anchored score · Vals RSI Index · LM training ↗

0.02 reference score · Kimi K2.5 (max) · Harness engineering · reference-anchored score · Vals RSI Index · Harness engineering ↗

0.00 reference score · Kimi K2.5 (max) · Post-training · reference-anchored score · Vals RSI Index · Post-training ↗

Fortress ↗41.09%
5.7874.39

kimi-k2.5 · Scale Labs · Risk elicitation score ↗

VISTA ↗41.86%
15.1854.65

kimi-k2.5 · Scale Labs · Score ↗

OmniDocBench ↗84.53 points
78.4496.91

Kimi K2.5 · OmniDocBench · Overall parsing score · end-to-end evaluation ↗

ALE-Bench ↗821.65 Performance points
137.782951.32

kimi-k2.5 · Average contest performance ↗

Closest models by score

Similar positions on shared benchmark leaderboards. Smaller gaps are closer.

18.8 percentile-point gap

21 of 30 shared benchmark comparisons

How similarity is measured

Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.

Filter comparisons to this model →

Reports featuring this model

Kimi K2.6 · earlier-2026 baseline · Kimi →

Qwen3.5 · earlier-2026 baseline · Qwen →

Sources

Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.

  1. Arena leaderboard — WebDev evaluation ↗

    Arena

  2. SciCode Benchmark | Artificial Analysis ↗

    Artificial Analysis

  3. TasteVal: Measuring the Experimental Research Taste of AI Systems Against Human Experts — published result table ↗

    Jaffe, Oliver, Sherburn, Dane · 2026

  4. Ale Bench — methodology and published results ↗

    Epoch AI · 2026

  5. ARC-AGI-1 ↗

    Epoch AI

  6. ARC-AGI-2 ↗

    Epoch AI

  7. CL-bench ↗

    Epoch AI

  8. CL-bench Life ↗

    Epoch AI

  9. CritPt v1 ↗

    Epoch AI

  10. Humanity's Last Exam ↗

    Epoch AI

  11. SimpleQA Verified ↗

    Epoch AI

  12. SimpleBench ↗

    Epoch AI

  13. SWE-bench Verified ↗

    Epoch AI

  14. Vending Bench 2 — methodology and published results ↗

    Epoch AI · 2026

  15. AI benchmarks & capabilities ↗

    Epoch AI

  16. Fiction.liveBench — March 14, 2025 benchmark release ↗

    Fiction.live · 2025

  17. WeirdML ↗

    Håvard Tveit Ihle

  18. Qwen/Qwen3.5-397B-A17B · Hugging Face ↗

    Qwen Team

  19. moonshotai/Kimi-K2.6 · Hugging Face ↗

    Moonshot AI

  20. EnigmaEval ↗

    Scale Labs / SEAL

  21. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  22. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  23. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  24. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  25. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  26. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  27. SWE Atlas - Refactoring — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  28. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  29. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  30. OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments ↗

    Xie, Tianbao, Zhang, Danyang, Chen, Jixuan, Li, Xiaochuan, Zhao, Siheng, Cao, Ruisheng et al. · 2024

    All 17 authors

    Xie, Tianbao, Zhang, Danyang, Chen, Jixuan, Li, Xiaochuan, Zhao, Siheng, Cao, Ruisheng, Hua, Toh Jing, Cheng, Zhoujun, Shin, Dongchan, Lei, Fangyu, Liu, Yitao, Xu, Yiheng, Zhou, Shuyan, Savarese, Silvio, Xiong, Caiming, Zhong, Victor, Yu, Tao

  31. OmniDocBench: Benchmarking Diverse PDF Document Parsing with Comprehensive Annotations — published result table ↗

    Linke Ouyang, Yuan Qu, Hongbin Zhou, Jiawei Zhu, Rui Zhang, Qunshu Lin et al. · 2024

    All 20 authors

    Linke Ouyang, Yuan Qu, Hongbin Zhou, Jiawei Zhu, Rui Zhang, Qunshu Lin, Bin Wang, Zhiyuan Zhao, Man Jiang, Xiaomeng Zhao, Jin Shi, Fan Wu, Pei Chu, Minghao Liu, Zhenxiang Li, Chao Xu, Bo Zhang, Botian Shi, Zhongying Tu, Conghui He

  32. Terminal-Bench 2.0 leaderboard ↗

    Terminal-Bench Team

  33. AIME — evaluation methodology and results ↗

    Vals AI · 2026

  34. CaseLaw — evaluation methodology and results ↗

    Vals AI · 2026

  35. Code Migration v1 — evaluation methodology and results ↗

    Vals AI · 2026

  36. CorpFin — evaluation methodology and results ↗

    Vals AI · 2026

  37. Excel Modeling Benchmark v1 — evaluation methodology and results ↗

    Vals AI · 2026

  38. Finance Agent v2 — evaluation methodology and results ↗

    Vals AI · 2026

  39. GPQA Diamond — evaluation methodology and results ↗

    Vals AI · 2026

  40. Harvey's Legal Agent Benchmark v1 — evaluation methodology and results ↗

    Vals AI · 2026

  41. LiveCodeBench — evaluation methodology and results ↗

    Vals AI · 2026

  42. Legal Research Bench — evaluation methodology and results ↗

    Vals AI · 2026

  43. MedCode — evaluation methodology and results ↗

    Vals AI · 2026

  44. MedQA — evaluation methodology and results ↗

    Vals AI · 2026

  45. MedScribe — evaluation methodology and results ↗

    Vals AI · 2026

  46. MMLU Pro — evaluation methodology and results ↗

    Vals AI · 2026

  47. MMMU Pro — evaluation methodology and results ↗

    Vals AI · 2026

  48. MortgageTax — evaluation methodology and results ↗

    Vals AI · 2026

  49. Vals RSI Index — evaluation methodology and results ↗

    Vals AI · 2026

  50. SAGE — evaluation methodology and results ↗

    Vals AI · 2026

  51. SWE-bench Verified — evaluation methodology and results ↗

    Vals AI · 2026

  52. TaxEval — evaluation methodology and results ↗

    Vals AI · 2026

  53. Vibe Code Bench — evaluation methodology and results ↗

    Vals AI · 2026

Names used in source tables

K2.5-1T-A32B · Kimi K2.5 · Kimi K2.5 (Mini-SWE-Agent) · Kimi K2.5 (Mini-SWE-agent) · Kimi K2.5 (OpenHands) · Kimi K2.5 (max) · Kimi K2.5 (max, 96k output) · Kimi K2.5 (thinking) · Kimi-K2.5 (Mini-SWE-Agent) · kimi-k2.5