THE FORGE RANKINGS·2026-08-30

Kimi K2.7 Code

Moonshot
23
RANK of 37 ranked
FERROX INDEX61.5 weighted mean of category positions
GRADEC below the field median
EVIDENCE8 / 32 benchmarks measured
QUALITYFULL every category rests on independent benchmarks

Not separable from Claude Sonnet 4.6 and Claude 4.5 Opus at the measured 1.78 point band. The order is printed; the gap is not claimed.

No measurement in Preference. That weight was redistributed, which is the same as assuming this model would have scored its own average there. It is an assumption, not a measurement.

Compared to what

Each rail is one classification. Every ranked model measured on it is a tick; this model is the orange marker. A hatched rail means this model has no measurement in that classification and its weight was redistributed.

Agents

33of 37 measured
25.0
13.1median 38.963.0

Coding

36of 37 measured
28.6
22.4median 50.070.7

Reasoning

2of 37 measured
91.7
18.1median 62.893.0

Preference

no qualifying measurement
53.1median 71.882.5

Every measurement, and the field behind it

One card per benchmark this model has been measured on. The rail shows where it sits against every other ranked model measured on the same benchmark. Where the evaluator keys on a harness, the harness is named, because the same model scores differently under a different one.

GPQA Diamond (Epoch's own run)

16/36
87.9% best 94.1% · Gemini 3.1 Pro Pr…
Epoch AI CC-BY-4.0

OTIS Mock AIME 2024-2025

9/36
95.6% best 99.7% · Claude Fable 5
Epoch AI CC-BY-4.0

ALE-Bench

27/35
886 best 2177 · GPT-5.6 Sol
ALE-Bench (Sakana AI with AtCoder) CC-BY-4.0

APEX Agents

16/28
27.6% best 45.0% · Claude Fable 5
APEX CC-BY-4.0

LMArena Agent

17/21
0.017 best 0.127 · Claude Opus 5
LMArena CC-BY-4.0

DeepSWE

15/17
30.5% best 72.8% · Claude Opus 5
deepswe.datacurve.ai CC-BY-4.0 mini-swe-agent

FrontierCode

13/17
30.1% best 53.5% · Claude Fable 5
cognition.com CC-BY-4.0 mini-swe-agent none

LMArena WebDev

10/16
1473 best 1626 · Claude Fable 5
LMArena CC-BY-4.0

Not measured

24/32
  • ARC-AGI-2
  • FrontierMath
  • SWE-bench Verified (Epoch's own run)
  • LMArena Text (style-controlled)
  • Humanity's Last Exam
  • Terminal-Bench 2
  • METR time horizons
  • Terminal-Bench 4.0
  • Aider Polyglot
  • OSWorld
  • Cybench
  • BrowseComp
  • IMOAnswerBench
  • SWE-bench Multilingual
  • AIME
  • CyberGym
  • HMMT
  • MCP-Atlas
  • SWE-bench Pro
  • Tau2-Bench Airline
  • Tau2-Bench Banking
  • Tau2-Bench Retail
  • Tau2-Bench Telecom
  • Tool-Decathlon
A missing benchmark is not a zero and is never scored as one.

Check it yourself

Every number on this page, with who measured it, the interval they published, the harness it was run under, and where to go and read it. Nothing here was measured by Ferrox Labs.

BenchmarkClassPublishedInterval (normalised) Measured byLicenceHarness / effort
APEX Agents agents 27.6% 21.1 to 34.1 APEX CC-BY-4.0 source
LMArena Agent agents 0.017 14.8 to 29.9 LMArena CC-BY-4.0 source
ALE-Bench coding 886 not published ALE-Bench (Sakana AI with AtCoder) CC-BY-4.0 source
DeepSWE coding 30.5% 30.0 to 31.0 deepswe.datacurve.ai CC-BY-4.0 mini-swe-agent source
FrontierCode coding 30.1% not published cognition.com CC-BY-4.0 mini-swe-agent / none source
LMArena WebDev preference 1473 57.9 to 60.3 LMArena CC-BY-4.0 source
GPQA Diamond (Epoch's own run) reasoning 87.9% 83.3 to 92.4 Epoch AI CC-BY-4.0 source
OTIS Mock AIME 2024-2025 reasoning 95.6% 89.5 to 100.0 Epoch AI CC-BY-4.0 source

Configurations rolled up: 3. Rule: median observed configuration per benchmark (lower median, always a real measurement). Harnesses seen: mini-swe-agent.

Snapshot 2026-08-30-62e4e85f043b, manifest dc46131315046f1e. Grades are positional: position in the measured field, as a percentile of rank among ranked entries, n=37.