Inference
Epoch AI

Furniture Assembly

Evaluated by Epoch AI on their own harness.

ModelsModels with a result.19
Top resultBest result on this test.80.0%GPT-6 Astra
Top-3 spreadPoints from first to third.19.2 pts
YearYear of release.N/A

Result and price

020406080100$1$3$10$30$100RESULTOUTPUT PRICE PER 1M TOKENS · LOG SCALE

Results

19 results
#ModelResultThe score from the source.
01GPT-6 Astramax effort80.0% ±4.9
02Claude Fable 5.1max effort70.0% ±5.7
03Claude Opus 5max effort60.8% ±6.2
04GPT-5.6 Solmax effort56.7% ±6.3
05GPT-5.6 Terramax effort54.2% ±6.4
06GPT-5.5xhigh effort44.2% ±6.4
07Claude Opus 4.8max effort42.5% ±6.4
07GPT-5.6 Lunamax effort42.5% ±6.4
09GPT-5.2xhigh effort38.3% ±6.3
10GPT-5.4xhigh effort37.5% ±6.2
11Claude Fable 5max effort35.8% ±6.2
12Claude Opus 4.7max effort33.3% ±6.1
13Gemini 3.8 Flashhigh effort31.7% ±5.9
14Claude Opus 4.6max effort28.3% ±5.9
14Claude Opus 4.528.3% ±5.9
16Gemini 3.7 Flashhigh effort26.7% ±5.6
16Gemini 3.1 Prohigh effort26.7% ±5.6
18Gemini 3.6 Flashhigh effort23.3% ±5.4
19Qwen3.8 Max (0902)xhigh effort20.0% ±5.2