Atlas

Benchmarks

← All benchmarks

RadLE 2.0 Accuracy Index (RadLE-A)

Multimodal

RadLE-A is raw diagnostic accuracy across all 200 RadLE 2.0 cases, independent of confidence. An explicit 'I don't know' is counted as not correct.

Top models (higher is better)

ModelScore
Human Expert Baseline (RadLE 2.0)38.5
Gemini 3.1 Pro Preview32.0
GPT-5.6 Sol Pro29.0
Muse Spark 1.128.5
Claude Fable 525.0
Qwen3.7-Plus16.5
GLM 5V Turbo16.0
Grok 4.511.5
Gemma 4 31B (RadLE 2.0 label)8.5
Llama 4 Maverick (RadLE 2.0 label)7.5
MiniMax M35.0
Lingshu-32B2.5
OctoMed-7B2.0
InternVL3.5-8B0.5
MedGemma 1.5 4B IT0.5
Mistral Large 3 2512 (RadLE 2.0 label)0.5
Nemotron 3 Nano Omni 30B A3B0.5
Loading Atlas data…