Atlas

Benchmarks

GDP.pdf (November 2026 rescoring)

Surge's GDP.pdf board as rescored between its November 6 and November 18, 2026 site builds, on the 100 held-out professional PDF tasks. The page has no changelog and names no metric, judge or run count. Surge's GitHub harness, unchanged since August 14, still describes five no-tools Inspect runs per task that pass only when Gemini 3.5 Flash judges every rubric criterion met, and the launch post still says no frontier model cleared 25%. Every published value is a multiple of 0.2, which fits five runs on each of the 100 tasks. Kept apart from the original board because every run it kept rose, by 16.2 to 26.0 points (Gemini 4 Argon High, 28.4 to 53.6; Grok 4.7 xHigh, 22.8 to 48.8).

Models
24
Top score
56.4
Leader
Gemini 4 Carbon
Updated
—
Resolution
~3 mo
Useful
Saturation
2028–2031
Projected
Signal
50thpercentile
141 benchmarks from 0.8 to 2.9196 benchmarks from 2.9 to 5.0131 benchmarks from 5.0 to 7.1105 benchmarks from 7.1 to 9.282 benchmarks from 9.2 to 11.365 benchmarks from 11.3 to 13.327 benchmarks from 13.3 to 15.421 benchmarks from 15.4 to 17.58 benchmarks from 17.5 to 19.68 benchmarks from 19.6 to 21.76 benchmarks from 21.7 to 23.84 benchmarks from 23.8 to 25.95 benchmarks from 25.9 to 28.03 benchmarks from 28.0 to 30.12 benchmarks from 30.1 to 32.14 benchmarks from 32.1 to 34.21 benchmark from 34.2 to 36.33 benchmarks from 36.3 to 38.41 benchmark from 38.4 to 40.52 benchmarks from 40.5 to 42.61 benchmark from 42.6 to 44.71 benchmark from 44.7 to 46.81 benchmark from 48.9 to 50.9

413th strongest of 818

Signal-to-noise +5.85

Difficulty
82ndpercentile
1 benchmark from 5.8 to 14.71 benchmark from 23.5 to 32.42 benchmarks from 32.4 to 41.21 benchmark from 41.2 to 50.12 benchmarks from 50.1 to 58.99 benchmarks from 58.9 to 67.88 benchmarks from 67.8 to 76.718 benchmarks from 76.7 to 85.523 benchmarks from 85.5 to 94.441 benchmarks from 94.4 to 103.232 benchmarks from 103.2 to 112.141 benchmarks from 112.1 to 121.049 benchmarks from 121.0 to 129.875 benchmarks from 129.8 to 138.760 benchmarks from 138.7 to 147.5122 benchmarks from 147.5 to 156.4135 benchmarks from 156.4 to 165.3124 benchmarks from 165.3 to 174.139 benchmarks from 174.1 to 183.021 benchmarks from 183.0 to 191.88 benchmarks from 191.8 to 200.72 benchmarks from 200.7 to 209.54 benchmarks from 209.5 to 218.4

148th hardest of 818

Capability needed 167.9

Organizations
24 dated models · 6 frontier modelsClosed weightsOpen weights
Data: Surge AIatlas.village.dev

GDP.pdf (November 2026 rescoring) leaderboard

#ModelSettingScoreSource
1Gemini 4 Carbonhigh
56.4
2Muse Wondermax
54.8
3Gemini 4 Hydrogenx-high
54.2
4Gemini 4 Argonhigh
53.6
5Gemini 4 Neonhigh
53.4
6Muse Spark 1.3max
53.0
7Claude Fable 5.2max
51.8
8GPT-6.1 Solmax
51.6
9Claude Sonnet 5.5max
51.2
10GPT-6 Astramax
50.4
11Claude Fable 5.1max
49.4
12GPT-6.2 Astramax
49.2
13Muse Embermax
49.2
14Claude Opus 5.5max
48.8
15GPT-6.2 Solmax
48.8
16Grok 4.7x-high
48.8
17Kimi K3.1 Promax
46.6
18Gemini 3.8 Flashhigh
44.2
19Hy4high
43.6
20Inkling 2 Small0.99
43.6
21Qwen4-Maxx-high
43.6
22Kimi K3max
39.8
23GLM-5.3max
36.6
24MiniMax M3.1 Flashmax
27.0

Related benchmarks

Loading Atlas data…