Atlas

Models

← All models

Apodex 1.1

Apodex · 2026-08-24

Apodex's proprietary 1.1 reasoning model, released in its online workbench on August 24, 2026 and evaluated by Artificial Analysis with text and image inputs. This entry denotes the hosted model, separate from the smaller locally deployable 1.1 mini release and the FrontierAgent harness.

Benchmark scores

BenchmarkScore
AA-Briefcase v1.1 Elo1013
AA-LCR v1.179.3
AA-Omniscience Index-21.9
APEX-Agents-AA (avg@3)31.2
Artificial Analysis Intelligence Index v4.3.226.4
Artificial Analysis Omniscience Accuracy31.7
Artificial Analysis Omniscience Hallucination Rate78.4
Artificial Analysis Progress Index v1292
AutomationBench-AA 1.0.628.2
Capability153.4
CritPt (avg@5)4.6
GDP.pdf (Artificial Analysis, All-pass)11.2
GDP.pdf (Artificial Analysis, Mean Criteria)63.3
GDPval-AA v21273
GDPval-AA v2.11196
GPQA Diamond (Artificial Analysis avg@5)86.4
Humanity's Last Exam (Text-Only)34.1
Humanity's Last Exam Diamond11.6
MMMU Pro79.2
SciCode 1.0.1 (Artificial Analysis)45.5
Terminal-Bench 2.169.7
Terminal-Bench 4.0 (Artificial Analysis)0.0
𝜏³-Banking25.2
Loading Atlas data…