THE FORGE RANKINGS·2026-08-30

Claude Sonnet 5

Anthropic
13
RANK of 37 ranked
FERROX INDEX74.2 weighted mean of category positions
GRADEB above the field median
EVIDENCE8 / 32 benchmarks measured
QUALITYFULL every category rests on independent benchmarks

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

31of 37 measured
29.0
13.1median 38.963.0

Coding

27of 37 measured
44.3
22.4median 50.070.7

Reasoning

10of 37 measured
80.2
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.

ALE-Bench

10/35
1463 best 2177 · GPT-5.6 Sol
ALE-Bench (Sakana AI with AtCoder) CC-BY-4.0 high

APEX Agents

13/28
32.5% best 45.0% · Claude Fable 5
APEX CC-BY-4.0

LMArena Agent

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

FrontierCode

8/17
42.7% best 53.5% · Claude Fable 5
cognition.com CC-BY-4.0 claude-code xhigh

Terminal-Bench 4.0

9/10
12.4% best 51.8% · Claude Opus 5
Terminal-Bench Apache-2.0 claude-code max

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
  • LMArena WebDev
  • METR time horizons
  • 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 32.5% 25.3 to 39.8 APEX CC-BY-4.0 source
LMArena Agent agents 0.077 34.7 to 49.7 LMArena CC-BY-4.0 source
Terminal-Bench 4.0 agents 12.4% 9.4 to 15.5 Terminal-Bench Apache-2.0 claude-code / max source
ALE-Bench coding 1463 not published ALE-Bench (Sakana AI with AtCoder) CC-BY-4.0 / high source
DeepSWE coding 48.2% 43.7 to 52.7 deepswe.datacurve.ai CC-BY-4.0 mini-swe-agent / high source
FrontierCode coding 42.7% not published cognition.com CC-BY-4.0 claude-code / xhigh source
GPQA Diamond (Epoch's own run) reasoning 80.3% 74.8 to 85.9 Epoch AI CC-BY-4.0 / max source
OTIS Mock AIME 2024-2025 reasoning 80.0% 68.2 to 91.8 Epoch AI CC-BY-4.0 / max source

Configurations rolled up: 11. Rule: median observed configuration per benchmark (lower median, always a real measurement). Harnesses seen: claude-code, 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.