Evaluation sources/MedCode

ObservingProfessional work

MedCode

Vals AI · Public model evaluation results from MedCode.

Official evaluation

Source description

This source measures concrete delivery performance in Professional work; raw scores are shown separately from the normalized capability score.

Limitation: Different versions, reasoning settings, and test sets may change scores; read them with the data date.

How it is used

Participating sources affect capability scores; reference, cross-reference, and observing sources are for comparison and verification only.

Model results from this source


Claude Opus 5	63.57	63.6	2026-11-05
Gemini 4 Argon	58.8	58.8	2026-11-05
Claude Fable 5	56.07	56.1	2026-11-05
Claude Opus 4.7	54.86	54.9	2026-11-05
Claude Fable 5.1	53.51	53.5	2026-11-05
Claude Opus 4.8	53.22	53.2	2026-11-05
Gemini 3.6 Flash	53.15	53.1	2026-11-05
Claude Sonnet 5.5	52.92	52.9	2026-11-05
Claude Haiku 5.5	51.54	51.5	2026-11-05
Gemini 2.5 Pro	50.59	50.6	2026-11-05
Claude Opus 5.5	49.8	49.8	2026-11-05
GPT-5.2	49.75	49.8	2026-11-05
GPT-5	49.63	49.6	2026-11-05
Grok 4.7	49.55	49.5	2026-11-05
Kimi K3	49.36	49.4	2026-11-05
Muse Spark 1.2	49.35	49.4	2026-11-05
Claude Opus 4.5	49.16	49.2	2026-11-05
Claude Opus 4.6	49.13	49.1	2026-11-05
GPT-5.5	49.1	49.1	2026-11-05
GPT-6.1 Sol	48.84	48.8	2026-11-05
GPT-6 Astra	48.49	48.5	2026-11-05
Gemini 3.8 Flash	48.13	48.1	2026-11-05
Gemini 3.1 Flash Lite Preview	47.6	47.6	2026-11-05
Claude Sonnet 5	47.54	47.5	2026-11-05

Page 1 of 3

Frequently asked questions