← All benchmarks
RCM-Bench-1: Encounter billing
Health & Medicine · 2026-10-02
Pass rate over three attempts on each of the 11 RCM-Bench-1 scenarios for adjusting balances, applying write-offs and managing patient responsibility.
Benchmark source: Commure
Top models (higher is better)
| Model | Score | Source |
|---|
| Opus 4.6 | 100.0 | RCM-Bench-1 results: Claude Opus 4.6, Encounter billing · Commure |
| Claude Opus 5.5 | 100.0 | RCM-Bench-1 results: Claude Opus 5.5, Encounter billing · Commure |
| DeepSeek V4.1 Flash | 97.0 | RCM-Bench-1 results: DeepSeek V4.1 Flash, Encounter billing · Commure |
| Sonnet 4.6 | 93.9 | RCM-Bench-1 results: Claude Sonnet 4.6, Encounter billing · Commure |
| Claude Sonnet 5.5 | 93.9 | RCM-Bench-1 results: Claude Sonnet 5.5, Encounter billing · Commure |
| GPT-5.6 Sol | 93.9 | RCM-Bench-1 results: GPT-5.6 Sol, Encounter billing · Commure |
| Grok 4.7 | 93.9 | RCM-Bench-1 results: Grok 4.7, Encounter billing · Commure |
| GPT-6.1 Sol | 90.9 | RCM-Bench-1 results: GPT-6.1 Sol, Encounter billing · Commure |
| Claude Opus 5 | 87.9 | RCM-Bench-1 results: Claude Opus 5, Encounter billing · Commure |
| Sonnet 5 | 87.9 | RCM-Bench-1 results: Claude Sonnet 5, Encounter billing · Commure |
| GLM-5.3 | 87.9 | RCM-Bench-1 results: GLM-5.3, Encounter billing · Commure |
| GPT-5.6 Terra | 87.9 | RCM-Bench-1 results: GPT-5.6 Terra, Encounter billing · Commure |
| GPT-5 Mini | 87.9 | RCM-Bench-1 results: GPT-5 mini, Encounter billing · Commure |
| Claude Haiku 5.5 | 84.8 | RCM-Bench-1 results: Claude Haiku 5.5, Encounter billing · Commure |
| GPT-6 Luna | 84.8 | RCM-Bench-1 results: GPT-6 Luna, Encounter billing · Commure |
| Haiku 4.5 | 81.8 | RCM-Bench-1 results: Claude Haiku 4.5, Encounter billing · Commure |
| Gemini 3.8 Flash | 81.8 | RCM-Bench-1 results: Gemini 3.8 Flash, Encounter billing · Commure |
| Gemini 4 Argon | 78.8 | RCM-Bench-1 results: Gemini 4 Argon, Encounter billing · Commure |
| GPT-6 Sol | 78.8 | RCM-Bench-1 results: GPT-6 Sol, Encounter billing · Commure |
| GPT-5.6 Luna | 66.7 | RCM-Bench-1 results: GPT-5.6 Luna, Encounter billing · Commure |