← All benchmarks
RCM-Bench-1: Eligibility checks
Health & Medicine · 2026-10-02
Pass rate over three attempts on each of the 5 RCM-Bench-1 scenarios for checking coverage, rechecking updates and handling inconclusive results.
Benchmark source: Commure
Top models (higher is better)
| Model | Score | Source |
|---|
| Haiku 4.5 | 100.0 | RCM-Bench-1 results: Claude Haiku 4.5, Eligibility checks · Commure |
| Opus 4.6 | 100.0 | RCM-Bench-1 results: Claude Opus 4.6, Eligibility checks · Commure |
| Claude Opus 5 | 100.0 | RCM-Bench-1 results: Claude Opus 5, Eligibility checks · Commure |
| Sonnet 4.6 | 100.0 | RCM-Bench-1 results: Claude Sonnet 4.6, Eligibility checks · Commure |
| Claude Sonnet 5.5 | 100.0 | RCM-Bench-1 results: Claude Sonnet 5.5, Eligibility checks · Commure |
| Sonnet 5 | 100.0 | RCM-Bench-1 results: Claude Sonnet 5, Eligibility checks · Commure |
| DeepSeek V4.1 Flash | 100.0 | RCM-Bench-1 results: DeepSeek V4.1 Flash, Eligibility checks · Commure |
| Gemini 3.8 Flash | 100.0 | RCM-Bench-1 results: Gemini 3.8 Flash, Eligibility checks · Commure |
| Gemini 4 Argon | 100.0 | RCM-Bench-1 results: Gemini 4 Argon, Eligibility checks · Commure |
| GLM-5.3 | 100.0 | RCM-Bench-1 results: GLM-5.3, Eligibility checks · Commure |
| GPT-5.6 Sol | 100.0 | RCM-Bench-1 results: GPT-5.6 Sol, Eligibility checks · Commure |
| GPT-6 Sol | 100.0 | RCM-Bench-1 results: GPT-6 Sol, Eligibility checks · Commure |
| Claude Haiku 5.5 | 93.3 | RCM-Bench-1 results: Claude Haiku 5.5, Eligibility checks · Commure |
| GPT-5.6 Terra | 93.3 | RCM-Bench-1 results: GPT-5.6 Terra, Eligibility checks · Commure |
| Grok 4.7 | 93.3 | RCM-Bench-1 results: Grok 4.7, Eligibility checks · Commure |
| Claude Opus 5.5 | 86.7 | RCM-Bench-1 results: Claude Opus 5.5, Eligibility checks · Commure |
| GPT-5 Mini | 86.7 | RCM-Bench-1 results: GPT-5 mini, Eligibility checks · Commure |
| GPT-6.1 Sol | 86.7 | RCM-Bench-1 results: GPT-6.1 Sol, Eligibility checks · Commure |
| GPT-6 Luna | 86.7 | RCM-Bench-1 results: GPT-6 Luna, Eligibility checks · Commure |
| GPT-5.6 Luna | 73.3 | RCM-Bench-1 results: GPT-5.6 Luna, Eligibility checks · Commure |