# Kimi K3High > Source: Benchize (https://benchize.com) — https://benchize.com/models/kimi-k3-high. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/models/kimi-k3-high. Moonshot AI Canonical URL: https://benchize.com/models/kimi-k3-high Last updated: Thu, 22 Oct 2026 07:10:12 GMT Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations. Moonshot AI ## Benchmarks | Benchmark | Score | ± | Rank | Cost | Source | | | --- | --- | --- | --- | --- | --- | --- | | [ARC-AGI-1](https://benchize.com/benchmarks/arc-prize%3Av1_Semi_Private.md) | 86.67% | — | 96 / 245 | 0.48usd per task | [arc prize · v1](https://arcprize.org/media/data/leaderboard/v1.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.8666666666666666 \| score_fraction \| v1_Semi_Private \| Higher is better \| \| `cost_per_task` \| 0.47977101000000005 \| usd_per_task \| v1_Semi_Private \| Lower is better \| ### Configuration evaluationSplit v1_Semi_Private harness Not reported modelGroup moonshot-kimi-k3 modelType CoT Reasoning effort high sourceModelVersion moonshot-kimi-k3-high thinkingEnabled true ### Source [Original source](https://arcprize.org/media/data/leaderboard/v1.json) Source arc prize Original model label Kimi K3 (High) Source version v1 Observed 2026-10-22T05:20:04.066Z Evidence SHA-256 `2bf68d22dafba37047d07ad833ebc0e8199c9aa2f9eeaed50c75c1b704462836` Source locator `/evaluations/183` | | | | | | | | [ARC-AGI-2](https://benchize.com/benchmarks/arc-prize%3Av2_Semi_Private.md) | 55% | — | 93 / 243 | 0.95usd per task | [arc prize · v2](https://arcprize.org/media/data/leaderboard/v2.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.55 \| score_fraction \| v2_Semi_Private \| Higher is better \| \| `cost_per_task` \| 0.9472229249999998 \| usd_per_task \| v2_Semi_Private \| Lower is better \| ### Configuration evaluationSplit v2_Semi_Private harness Not reported modelGroup moonshot-kimi-k3 modelType CoT Reasoning effort high sourceModelVersion moonshot-kimi-k3-high thinkingEnabled true ### Source [Original source](https://arcprize.org/media/data/leaderboard/v2.json) Source arc prize Original model label Kimi K3 (High) Source version v2 Observed 2026-10-22T05:20:04.306Z Evidence SHA-256 `4351eb696e78113b0459e9bb5f0258318b905efef34f473dcea51ce641b0d94d` Source locator `/evaluations/184` | | | | | | | | [ToneBench Scriptwriting](https://benchize.com/benchmarks/tonebench%3Ascripts.md) | 1,965.4 | Range 1,916.3–2,008.5 | 39 / 126 | — | [tonebench](https://benchmark.towardsai.com/data/leaderboard.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1965.4 \| rating \| reported \| Higher is better \| \| `overall` \| 80.35 \| percent \| reported \| Higher is better \| \| `anti_slop_numbers` \| 69.86 \| source_score \| rubric \| Higher is better \| \| `continuity_emotion` \| 81.06 \| source_score \| rubric \| Higher is better \| \| `cue_quality` \| 75.97 \| source_score \| rubric \| Higher is better \| \| `hook_strength` \| 83.24 \| source_score \| rubric \| Higher is better \| \| `length_adherence` \| 84.49 \| source_score \| rubric \| Higher is better \| \| `slop` \| 82 \| source_score \| rubric \| Higher is better \| \| `slop_score` \| 94.14 \| source_score \| rubric \| Higher is better \| \| `substance_accuracy` \| 82.38 \| source_score \| rubric \| Higher is better \| \| `tone_voice` \| 77.4 \| source_score \| rubric \| Higher is better \| \| `writing_quality` \| 77.94 \| source_score \| rubric \| Higher is better \| \| `youtube_best_practices` \| 79.89 \| source_score \| rubric \| Higher is better \| ### Configuration provider openrouter Reasoning effort high scoreScope 10_task_current sourceConfiguration kimi-k3-thinking-or thinkingEnabled true ### Reported statistics Uncertainty Reported interval: 1,916.3–2,008.5 interval [1916.3,2008.5] rank 39 runCount 5 ### Source [Original source](https://benchmark.towardsai.com/data/leaderboard.json) Source tonebench Original model label Kimi K3 (high) Evaluated 2026-10-08 Observed 2026-10-22T04:40:14.102Z Evidence SHA-256 `4264338337f418d3ded7ace250cad64fe12cacbd4b79a312bfaf0ae9268a01c3` Source locator `/rows/38` | | | | | | | | [WeirdML v3](https://benchize.com/benchmarks/weirdml%3Av3.md) | 0.08source score | Range 0.06–0.09 | 16 / 25 | — | [weirdml · b0120fe](https://htihle.github.io/assets/data/weirdml_v3.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.07544402727272727 \| source_score \| reported \| Higher is better \| \| `mean_final_best` \| 0.15409355909090908 \| source_score \| reported \| Higher is better \| ### Configuration budget {"kind":"equivalent_tokens","limit":50000000,"start":500000} harness kimi_code Reasoning effort high sourceConfiguration kimi-k3-fireworks-high-kimi-code ### Reported statistics Uncertainty Reported interval: 0.06–0.09 interval [0.062012804062472324,0.08904981193342236] intervalMethod {"assumptions":"Independent fresh runs; approximately normal scores; shared variance prior. Pool fitted within analysis mode, including incomplete models; pool estimation uncertainty is not propagated.","confidence":0.95,"draws":10000,"fixed_tasks":true,"method":"partially_pooled_variance","pool":{"degrees_of_freedom":1533,"fallback":false,"prior_df":4,"variance":0.0026383113392869583},"seed":20260915,"unit":"run"} runCount 72 taskCount 11 ### Source [Original source](https://htihle.github.io/assets/data/weirdml_v3.json) Source weirdml Original model label Kimi K3 (high, Fireworks) Source version b0120fe Observed 2026-10-22T04:50:05.075Z Evidence SHA-256 `82c3a8de423b23ca0af9761429299a2f21edb0b819e947fdc863f8b5ea4a49b8` Source locator `/rows/15` | | | | | | | | [NC Bench](https://benchize.com/benchmarks/nc-bench%3Aoverall.md) | 92.99% | — | 26 / 144 | — | [nc bench](https://www.nc-bench.com/models) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 92.99 \| percent \| reported \| Higher is better \| \| `tooling` \| 100 \| percent \| category \| Higher is better \| \| `creative_writing` \| 87.43 \| percent \| category \| Higher is better \| \| `language` \| 100 \| percent \| category \| Higher is better \| \| `utility` \| 96.19 \| percent \| category \| Higher is better \| \| `reasoning` \| 90.48 \| percent \| category \| Higher is better \| \| `text_editing` \| 97.42 \| percent \| category \| Higher is better \| \| `rule_following` \| 75.43 \| percent \| category \| Higher is better \| \| `hallucination` \| 96.97 \| percent \| category \| Higher is better \| ### Configuration Reasoning effort high Model configuration Kimi K3 (Reasoning, High) thinkingEnabled true ### Source [Original source](https://www.nc-bench.com/models) Source nc bench Original model label Kimi K3 (Reasoning, High) Observed 2026-10-22T06:10:03.426Z Evidence SHA-256 `ec261022458543624b08a136d01f5a2f0666cd52ffb49c64d6ca983e8543d1b5` Source locator `/rows/32` | | | | | | | --- Cite as: Kimi K3High — Benchize, https://benchize.com/models/kimi-k3-high (retrieved 2026-10-22)