SShogo Bench

PostTrainBench

The agent post-trains a small model end to end on a GPU budget; scored on the resulting model.

Agent Index weight
4%
in the Long-horizon autonomy pillar
Models tested
21
from the benchmark’s leaderboard and reports
Normalised for the index
0 → 0, 1 → 100
Official site

Results

#ModelScoreNormalisedHarnessCost/taskDate
1Claude Fable 5.1 Supported
Anthropic
44.6%Claude Code——
2Claude Fable 5.2 Supported
Anthropic
44.6%Claude Code——
3Claude Opus 5.5 Supported
Anthropic
43.8%Claude Code——
4GPT-6 Astra Supported
OpenAI
41.9%Codex CLI——
5Gemini 4 Carbon Supported
Google
41.0%OpenCode——
6Gemini 4 Argon Supported
Google
40.4%OpenCode——
7GPT-6.1 Sol Estimated
OpenAI
38.8%Codex CLI——
8Claude Fable 5 Supported
Anthropic
37.3%Claude Code——
9GLM-5.3-Flash Supported
Z.ai (Zhipu)
36.8%Claude Code——
10Claude Opus 5 Supported
Anthropic
35.7%Claude Code——
11GLM-5.3 Supported
Z.ai (Zhipu)
34.5%Claude Code——
12Kimi K3 Supported
Moonshot AI
31.5%Claude Code——
13GPT-5.6 Sol Supported
OpenAI
31.0%Codex CLI——
14Claude Opus 4.8 Supported
Anthropic
31.0%Claude Code——
15GLM-5.2 Supported
Z.ai (Zhipu)
30.4%Claude Code——
16Claude Opus 4.7 Supported
Anthropic
24.3%Claude Code——
17Muse Wonder Estimated
Meta
24.2%OpenCode——
18GPT-5.5 Supported
OpenAI
23.9%Codex CLI——
19Grok 4.5 Estimated
xAI
22.6%Cursor CLI——
20Gemini 3.1 Pro Preview Supported
Google
18.1%OpenCode——
21GPT-5.4 Supported
OpenAI
17.8%Codex CLI——

For models without a PostTrainBench result, the index estimates one from Epoch’s Capabilities Index using a logistic fit on 21 models (residual SD 0.3 on the logit scale). Estimates are labelled and widen the model’s range.