Kimi K2.5
Kimi K2.5 (Moonshot) was released on Jan 27, 2026. In the Epoch AI capability ranking (ECI), it ranks #67 out of 267 models, with 147.98. On Arena, which measures human preference in blind comparisons, it has 1,495 ELO points (#62). Against models of the same generation, its best result is on AIME 2024 and 2025 (OTIS): 92.2%, #14 of 61.
Where Kimi K2.5 stands out and where it falls behind, by the numbers
Kimi K2.5 has published results on 22 tests compiled by Epoch AI. The yardstick here is its generation: its position among the models released between May 2, 2025 and Apr 27, 2026 that took the same test. Against the whole archive, which includes 2023 models, a recent release lands near the top of almost everything, and that says nothing.
Where it stands out
Tests where it sits in the top third of its own generation.
Where it falls behind
Tests where it sits in the bottom half of its own generation.
See all 22 tests, one by one
| Test | Score | In its generation | Overall | Test record |
|---|---|---|---|---|
| GPQA Diamond | 87.6% | #20 / 61 | #57 / 173 | 96.4% Gemini 4 Argon |
| AIME 2024 and 2025 (OTIS) | 92.2% | #14 / 61 | #50 / 168 | 100.0% GPT-5.5 |
| CritPt | 3.1% | #18 / 43 (tied with 1) | #64 / 112 | 32.3% GPT-5.6 Sol |
| WeirdML | 45.6% | #35 / 53 | #59 / 112 | 93.6% GPT-6 Astra |
| SciCode | 49.0% | #10 / 38 | #52 / 104 | 66.9% Claude Opus 5.5 |
| ARC-AGI 2 | 11.8% | #20 / 38 | #48 / 83 | 95.0% GPT-6 Astra |
| ALE-Bench | 822 | #22 / 45 | #52 / 82 | 2,951 GPT-6 Astra |
| WebDev Arena | 1,436 | #16 / 41 | #54 / 82 | 1,820 Claude Opus 5.5 |
| ARC-AGI | 65.3% | #19 / 38 | #47 / 81 | 98.5% Claude Fable 5 |
| SimpleQA Verified | 34.3% | #19 / 33 | #53 / 81 | 76.0% Gemini 4 Argon |
| SimpleBench | 46.8% | #23 / 33 | #40 / 74 | 81.9% Claude Fable 5 |
| FrontierMath | 27.9% | #18 / 42 | #20 / 65 | 51.7% GPT-5.5 |
| Vending Bench 2 | 1,198 | #21 / 32 | #46 / 57 | 15,515 GPT-6 Astra |
| FrontierMath tier 4 | 4.2% | #19 / 40 | #22 / 48 | 39.6% GPT-5.5 Pro |
| Hle | 24.4% | #12 / 22 | #17 / 39 | 54.8% GPT-6 Astra |
| EnigmaEval | 3.4% | #16 / 19 | #24 / 36 | 39.3% Claude Fable 5 |
| Terminal-Bench | 43.2% | #18 / 33 | #18 / 33 | 84.7% GPT-5.5 |
| SWE-bench Verified | 73.8% | #14 / 25 (tied with 1) | #17 / 32 | 83.5% Claude Opus 4.7 |
| Fiction.liveBench | 78.1% | #5 / 17 | #6 / 31 | 100.0% o3 |
| Cl Bench | 19.3% | #9 / 18 | #9 / 19 | 27.9% GPT-5.4 |
| Cl Bench Life | 13.2% | #7 / 13 | #7 / 13 | 22.2% GPT-5.5 |
| Os World | 63.30 | #3 / 5 | #3 / 8 | 72.10 Claude Sonnet 4.6 |
For each test, the model's best effort level counts, as in the ranking. Source of the scores: Epoch AI, under CC BY 4.0.
What to compare Kimi K2.5 with
The models from other labs closest in capability and released in the same period, at most two per lab. Click any of them to see its page.
| Model | Lab | ECI | ELO | LiveBench | Release |
|---|---|---|---|---|---|
| GPT-5.4 Mini | OpenAI | 148.77 | 1,446 | 66.4 | Mar 17, 2026 |
| Inkling | Thinking Machines | 148.59 | 1,492 | 71.9 | Jul 15, 2026 |
| Kimi K2.5 | Moonshot | 147.98 | 1,495 | n/a | Jan 27, 2026 |
| Qwen 3.6 Plus | Alibaba | 147.49 | 1,476 | 68.9 | Mar 31, 2026 |
| o3-pro | OpenAI | 147.37 | n/a | n/a | Jun 10, 2025 |
| Qwen3.7-Plus | Alibaba | 147.30 | 1,510 | n/a | Jun 2, 2026 |
| MiniMax-M3 | MiniMax | 147.13 | 1,471 | 67.3 | Jun 1, 2026 |
Kimi K2.5 or Qwen 3.6 Plus?
Qwen 3.6 Plus is the model from another lab closest to Kimi K2.5 in capability. Across the 12 tests both took, this is the score:
| Test | Kimi K2.5 | Qwen 3.6 Plus |
|---|---|---|
| Vending Bench 2 | 1,198 | 5,115 |
| FrontierMath tier 4 | 4.2% | 8.3% |
| SimpleQA Verified | 34.3% | 44.1% |
| SWE-bench Verified | 73.8% | 57.9% |
| ALE-Bench | 822 | 670 |
| SciCode | 49.0% | 40.7% |
| CritPt | 3.1% | 2.9% |
| FrontierMath | 27.9% | 26.2% |
| Cl Bench | 19.3% | 20.3% |
| WebDev Arena | 1,436 | 1,461 |
Unweighted score: each test counts one point, whether math or coding. The dot marks who did better on each row.
Other models from Moonshot
The models from the same lab closest in time, newest first.
| Model | Release | ECI | ELO | LiveBench | Access |
|---|---|---|---|---|---|
| Kimi K3 | Jul 16, 2026 | 157.42 | 1,528 | 79.2 | Open weights (non-commercial) |
| Kimi K2.7 Code | Jun 12, 2026 | 149.87 | n/a | 68.4 | Open weights (unrestricted) |
| Kimi K2.6 | Apr 20, 2026 | 150.92 | 1,521 | 70.5 | Open weights (unrestricted) |
| Kimi K2.5 | Jan 27, 2026 | 147.98 | 1,495 | n/a | Open weights (unrestricted) |
| Kimi K2 Thinking | Nov 6, 2025 | 145.96 | n/a | n/a | Open weights (restricted use) |
| Kimi K2 Thinking Turbo | Nov 6, 2025 | 145.79 | 1,456 | n/a | Open weights (restricted use) |
| Kimi K2 Instruct | Jul 12, 2025 | 140.13 | n/a | n/a | Open weights (restricted use) |
Have you used Kimi K2.5? Rate it
Users' opinion, not measurement: the rating is kept apart from the ranking and does not move anyone's position. One vote per browser, no sign-up, and you can change your rating later.
What people ask about Kimi K2.5
Where does Kimi K2.5 rank among AI models?
In the ROO3 AI Benchmark, ordered by the Epoch AI ECI, Kimi K2.5 ranks #67 out of 267 models, with an ECI of 147.98. The ECI is a composite index that combines several hard benchmarks into a single scale. The position changes when a new model comes out, and the page is refreshed six times a day.
Is Kimi K2.5 better than Qwen 3.6 Plus?
Across the 12 tests both models took and Epoch AI compiled, Kimi K2.5 comes out ahead on 5, behind on 7 and tied on 0. It is a test-by-test count with no weighting: a win in math counts the same as a win in coding. To choose between the two, look at the tests that resemble your own use.
What is Kimi K2.5 best at?
Against models of the same generation (released up to 9 months before and up to 3 months after it), the best relative results of Kimi K2.5 are: AIME 2024 and 2025 (OTIS) (#14 of 61); SciCode (#10 of 38); Fiction.liveBench (#5 of 17). The yardstick is its position among the models of that period that took the same test.
Where do the numbers on this page come from?
From the same three sources as the ranking, all openly licensed: Epoch AI (ECI and individual tests, CC BY 4.0), Arena (human-preference ELO, CC BY 4.0) and LiveBench (overall average, CC BY-SA 4.0). ROO3 does not measure any model. When the page shows a figure published by the maker itself, it appears in the written analysis, labeled as the maker's, and never enters the ranking.
Where each number comes from
Epoch AI, "AI Benchmarking Hub". Published at epoch.ai. Data under the CC BY 4.0 licence, reordered by roo3.co.
The Arena (LMArena) leaderboard, from the public lmarena-ai/leaderboard-dataset on Hugging Face, under the CC BY 4.0 licence. We are not affiliated with the Arena.
LiveBench, by the LiveBench team, published at livebench.ai under the CC BY-SA 4.0 license. The overall average is recomputed by roo3.co with the same formula as the website (average of the categories), and those numbers stay under CC BY-SA 4.0. We are not affiliated with LiveBench.
← See the full ranking