| Claude Sonnet 5.5 | Anthropic | Model | 28 Sep 2026 | 63.27 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), PhysicianBench, RCM-Bench, DAYJOB Health, MLCR-AA |
| Claude Opus 5.5 | Anthropic | Model | 22 Sep 2026 | 62.88 of 12 boards | 8 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), PhysicianBench, RCM-Bench, DAYJOB Health, APEX-1 GP, MLCR-AA |
| Claude Opus 5 | Anthropic | Model | 24 Jul 2026 | 56.810 of 12 boards | 12 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), MAST (Medical AI Superintelligence Test), First, Do NOHARM (v2), PhysicianBench, HealthAgentBench, CHI-Bench, Bedside Bench, RCM-Bench, DAYJOB Health, APEX-1 GP |
| Claude Fable 5.1 | Anthropic | Model | 1 Sep 2026 | 56.86 of 12 boards | 6 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), PhysicianBench, DAYJOB Health, MLCR-AA |
| GPT-6 Astra | OpenAI | Model | 3 Sep 2026 | 56.17 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), Bedside Bench, DAYJOB Health, APEX-1 GP, MLCR-AA |
| Claude Haiku 5.5 | Anthropic | Model | 7 Oct 2026 | 55.54 of 12 boards | 4 / 14 | HealthBench Professional, PhysicianBench, RCM-Bench, MLCR-AA |
| GPT-6.1 Sol | OpenAI | Model | 29 Sep 2026 | 53.97 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), RCM-Bench, DAYJOB Health, APEX-1 GP, MLCR-AA |
| Muse Spark 1.1 | Meta | Model | 9 Jul 2026 | 52.73 of 12 boards | 3 / 14 | HealthBench Professional, MedScribe (Vals AI), First, Do NOHARM (v2) |
| Kimi K3.1 Pro | Moonshot AI | Model | 23 Oct 2026 | 52.41 of 12 boards | 1 / 14 | MLCR-AA |
| Gemini 4 Argon | Google | Model | 30 Sep 2026 | 51.86 of 12 boards | 6 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), Bedside Bench, RCM-Bench, DAYJOB Health |
| GPT-5.5 | OpenAI | Model | 23 Apr 2026 | 51.75 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), First, Do NOHARM (v2), PhysicianBench, HealthAgentBench, CHI-Bench |
| GPT-6.1 Terra | OpenAI | Model | 26 Oct 2026 | 51.72 of 12 boards | 2 / 14 | HealthBench Professional, MLCR-AA |
| Muse Wonder | Meta | Model | 12 Oct 2026 | 51.53 of 12 boards | 3 / 14 | HealthBench Professional, DAYJOB Health, MLCR-AA |
| GPT-5.6 Sol | OpenAI | Model | 9 Jul 2026 | 51.49 of 12 boards | 11 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM), MAST (Medical AI Superintelligence Test), First, Do NOHARM (v2), HealthAgentBench, CHI-Bench, Bedside Bench, RCM-Bench, APEX-1 GP |
| Claude Fable 5 | Anthropic | Model | 9 Jun 2026 | 50.86 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM), First, Do NOHARM (v2), CHI-Bench, DAYJOB Health |
| GPT-6 Luna | OpenAI | Model | 22 Sep 2026 | 50.66 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), PhysicianBench, Bedside Bench, RCM-Bench, MLCR-AA |
| Claude Opus 4.8 | Anthropic | Model | 28 May 2026 | 50.64 of 12 boards | 6 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM), HealthAgentBench, CHI-Bench |
| Kimi K3 | Moonshot AI | Model | 16 Jul 2026 | 50.47 of 12 boards | 8 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), MAST (Medical AI Superintelligence Test), First, Do NOHARM (v2), CHI-Bench, Bedside Bench, APEX-1 GP |
| Gemini 3.7 Flash | Google | Model | 13 Aug 2026 | 50.12 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Grok 4.7 | SpaceX AI | Model | 21 Sep 2026 | 50.07 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), RCM-Bench, DAYJOB Health, APEX-1 GP, MLCR-AA |
| Claude Sonnet 5 | Anthropic | Model | 30 Jun 2026 | 49.86 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), MAST (Medical AI Superintelligence Test), PhysicianBench, CHI-Bench, RCM-Bench |
| Gemini 3.5 Flash | Google | Model | 19 May 2026 | 49.72 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| GPT-6 Sol | OpenAI | Model | 22 Sep 2026 | 49.67 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), Bedside Bench, RCM-Bench, DAYJOB Health, MLCR-AA |
| Muse Spark | Meta | Model | 8 Apr 2026 | 49.64 of 12 boards | 4 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM) |
| Muse Wonder Full | Meta | Model | 12 Oct 2026 | 49.52 of 12 boards | 2 / 14 | HealthBench Professional, MLCR-AA |
| Gemini 3.6 Flash | Google | Model | 21 Jul 2026 | 49.13 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MAST (Medical AI Superintelligence Test) |
| GPT-5.6 Terra | OpenAI | Model | 9 Jul 2026 | 48.94 of 12 boards | 5 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), CHI-Bench, RCM-Bench |
| Gemini 3 Pro (11/25) | Google | Model | Not stated | 48.71 of 12 boards | 1 / 14 | MedCode (Vals AI) |
| GLM 5.3 Flash | Zhipu | Model | 26 Aug 2026 | 48.61 of 12 boards | 1 / 14 | MedScribe (Vals AI) |
| Gemini 3 Flash | Google | Model | 17 Dec 2025 | 48.52 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), CHI-Bench |
| DeepSeek V4.1 Flash 1017 | DEDeepSeek | Model | 17 Oct 2026 | 48.22 of 12 boards | 2 / 14 | Bedside Bench, MLCR-AA |
| Claude Opus 4.6 | Anthropic | Model | 5 Feb 2026 | 48.05 of 12 boards | 7 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM), PhysicianBench, HealthAgentBench, CHI-Bench, RCM-Bench |
| MiniMax M3 | MiniMax | Model | 1 Jun 2026 | 48.02 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Claude Opus 4.5 | Anthropic | Model | 24 Nov 2025 | 47.73 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM) |
| GLM 5.3 | Zhipu | Model | 18 Aug 2026 | 47.47 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), Bedside Bench, RCM-Bench, APEX-1 GP, MLCR-AA |
| Gemini 3.8 Flash | Google | Model | 2 Sep 2026 | 47.37 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), Bedside Bench, RCM-Bench, APEX-1 GP, MLCR-AA |
| Claude Opus 4.7 | Anthropic | Model | 16 Apr 2026 | 47.34 of 12 boards | 6 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), PhysicianBench, HealthAgentBench, CHI-Bench |
| GPT-5.6 Luna | OpenAI | Model | 9 Jul 2026 | 47.34 of 12 boards | 5 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), CHI-Bench, RCM-Bench |
| Muse Spark 1.3 | Meta | Model | 2 Sep 2026 | 47.31 of 12 boards | 1 / 14 | DAYJOB Health |
| Qwen3.8 Max | ALAlibaba | Model | 3 Aug 2026 | 47.04 of 12 boards | 4 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM), MLCR-AA |
| Grok 4.5 | SpaceX AI | Model | 8 Jul 2026 | 46.82 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Claude Sonnet 4.5 | Anthropic | Model | 29 Sep 2025 | 46.72 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Ember-1 | FAFireworks AI | Model | Not stated | 46.51 of 12 boards | 1 / 14 | Bedside Bench |
| o3 | OpenAI | Model | 16 Apr 2025 | 46.42 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Gemini 2.5 Flash | Google | Model | 17 Jun 2025 | 46.21 of 12 boards | 1 / 14 | MedScribe (Vals AI) |
| Qwen3.8 27B | ALAlibaba | Model | Not stated | 46.21 of 12 boards | 1 / 14 | MLCR-AA |
| GPT-5 | OpenAI | Model | 7 Aug 2025 | 46.14 of 12 boards | 4 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), First, Do NOHARM (v2) |
| Qwen3.7 Plus | ALAlibaba | Model | 31 May 2026 | 46.11 of 12 boards | 1 / 14 | MedXpertQA (MM) |
| GPT-5 mini | OpenAI | Model | 7 Aug 2025 | 45.93 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), RCM-Bench |
| Claude Opus 4.1 | Anthropic | Model | 5 Aug 2025 | 45.92 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Gemini 2.5 Pro | Google | Model | 17 Jun 2025 | 45.73 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), First, Do NOHARM (v2) |
| Qwen3.5 397B A17B | ALAlibaba | Model | 16 Feb 2026 | 45.73 of 12 boards | 3 / 14 | MedXpertQA (MM), MAST (Medical AI Superintelligence Test), First, Do NOHARM (v2) |
| MiMo V2.6 Flash | XIXiaomi | Model | 22 Sep 2026 | 45.62 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Grok 4.6 | SpaceX AI | Model | 12 Aug 2026 | 45.56 of 12 boards | 6 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), Bedside Bench, DAYJOB Health, APEX-1 GP |
| GPT-5.2 | OpenAI | Model | 11 Dec 2025 | 45.54 of 12 boards | 4 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM) |
| Gemini 3 Pro | Google | Model | 18 Nov 2025 | 45.52 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedXpertQA (MM) |
| DeepSeek V4.1 Flash | DEDeepSeek | Model | 10 Sep 2026 | 45.36 of 12 boards | 6 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), Bedside Bench, RCM-Bench, APEX-1 GP |
| GLM 5.2 | Zhipu | Model | 16 Jun 2026 | 45.22 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), CHI-Bench |
| DeepSeek V4 Flash | DEDeepSeek | Model | 24 Apr 2026 | 45.21 of 12 boards | 1 / 14 | MedScribe (Vals AI) |
| Muse Spark 1.2 | Meta | Model | 5 Aug 2026 | 44.93 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), DAYJOB Health |
| GPT-5.4 | OpenAI | Model | 5 Mar 2026 | 44.55 of 12 boards | 7 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM), PhysicianBench, HealthAgentBench, CHI-Bench |
| GPT-5.5 Instant | OpenAI | Model | 5 May 2026 | 44.42 of 12 boards | 2 / 14 | HealthBench Professional, MLCR-AA |
| DeepSeek V4 Flash 0731 | DEDeepSeek | Model | 31 Jul 2026 | 44.41 of 12 boards | 1 / 14 | MedCode (Vals AI) |
| Gemini 3.1 Flash Lite Preview | Google | Model | 3 Mar 2026 | 44.22 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Beam 1.0 | RAReflection AI | Model | Not stated | 44.21 of 12 boards | 1 / 14 | MLCR-AA |
| MiMo V2.6 Pro | XIXiaomi | Model | 22 Sep 2026 | 44.13 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MLCR-AA |
| GPT-5.4 nano | OpenAI | Model | 17 Mar 2026 | 44.12 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Inkling Small | TMThinking Machines | Model | 30 Jul 2026 | 44.12 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Gemini 2.5 Flash (7/17) | Google | Model | Not stated | 44.01 of 12 boards | 1 / 14 | MedCode (Vals AI) |
| Gemini 2.5 Flash Preview (9/25) | Google | Model | 25 Sep 2025 | 44.01 of 12 boards | 1 / 14 | MedCode (Vals AI) |
| Gemini 3.5 Flash Lite | Google | Model | 21 Jul 2026 | 43.92 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| MiniMax M3.1 Flash | MiniMax | Model | 19 Oct 2026 | 43.81 of 12 boards | 1 / 14 | MLCR-AA |
| Qwen 3.7 Max | ALAlibaba | Model | Not stated | 43.62 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Grok 4.3 | SpaceX AI | Model | Not stated | 43.53 of 12 boards | 4 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MAST (Medical AI Superintelligence Test), CHI-Bench |
| Grok 4 Fast (Reasoning) | SpaceX AI | Model | 19 Sep 2025 | 43.52 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Kimi K2.5 | Moonshot AI | Model | 27 Jan 2026 | 43.33 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM) |
| Nemotron 3 Ultra | NVIDIA | Model | 4 Jun 2026 | 43.31 of 12 boards | 1 / 14 | MedCode (Vals AI) |
| GPT-5.1 | OpenAI | Model | 12 Nov 2025 | 43.23 of 12 boards | 3 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI) |
| Grok 4 | SpaceX AI | Model | 9 Jul 2025 | 43.22 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Ling 3.1 Flash 1019 | AGAnt Group | Model | 27 Oct 2026 | 43.11 of 12 boards | 1 / 14 | MLCR-AA |
| Gemini 3.1 Pro | Google | Model | 19 Feb 2026 | 42.46 of 12 boards | 6 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM), MAST (Medical AI Superintelligence Test), First, Do NOHARM (v2), PhysicianBench |
| Nemotron 3.5 Super VL | NVIDIA | Model | 15 Oct 2026 | 42.41 of 12 boards | 1 / 14 | MLCR-AA |
| Hy4 Preview | TETencent | Model | 28 Aug 2026 | 42.33 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), DAYJOB Health |
| GLM 5.1 | Zhipu | Model | 7 Apr 2026 | 42.23 of 12 boards | 4 / 14 | MedScribe (Vals AI), MedCode (Vals AI), First, Do NOHARM (v2), CHI-Bench |
| Gemma 4 31B | Google | Model | 2 Apr 2026 | 42.21 of 12 boards | 1 / 14 | MedXpertQA (MM) |
| MiniMax-M2.1 | MiniMax | Model | Not stated | 42.12 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Muse Spark Open | Meta | Model | 15 Oct 2026 | 42.02 of 12 boards | 2 / 14 | Bedside Bench, MLCR-AA |
| K2 Horizon 375B A23B | IOInstitute of Foundation Models | Model | Not stated | 42.01 of 12 boards | 1 / 14 | MLCR-AA |
| DeepSeek V4 Pro | DEDeepSeek | Model | 24 Apr 2026 | 41.85 of 12 boards | 6 / 14 | MedScribe (Vals AI), MedCode (Vals AI), PhysicianBench, CHI-Bench, Bedside Bench, APEX-1 GP |
| DeepSeek R1 | DEDeepSeek | Model | 20 Jan 2025 | 41.61 of 12 boards | 1 / 14 | First, Do NOHARM (v2) |
| Claude Sonnet 4.6 | Anthropic | Model | 17 Feb 2026 | 41.53 of 12 boards | 5 / 14 | HealthBench Professional, PhysicianBench, HealthAgentBench, CHI-Bench, RCM-Bench |
| Gemini 2.5 Flash Lite (9/25) | Google | Model | 25 Sep 2025 | 41.51 of 12 boards | 1 / 14 | MedCode (Vals AI) |
| Ling 3.0 Flash | AGAnt Group | Model | Not stated | 41.42 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Gemma 4 26B A4B | Google | Model | 2 Apr 2026 | 41.01 of 12 boards | 1 / 14 | MedXpertQA (MM) |
| Claude Sonnet 4 | Anthropic | Model | 22 May 2025 | 40.92 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Inkling | TMThinking Machines | Model | 15 Jul 2026 | 40.63 of 12 boards | 4 / 14 | MedScribe (Vals AI), MedCode (Vals AI), CHI-Bench, MLCR-AA |
| Qwen 3 VL Plus | ALAlibaba | Model | Not stated | 40.52 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Qwen 3.8 27B | ALAlibaba | Model | Not stated | 40.52 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Grok 4 Fast (Non-Reasoning) | SpaceX AI | Model | 19 Sep 2025 | 40.32 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Kimi K2.6 | Moonshot AI | Model | 20 Apr 2026 | 39.94 of 12 boards | 5 / 14 | MedScribe (Vals AI), MedCode (Vals AI), First, Do NOHARM (v2), PhysicianBench, CHI-Bench |
| o4-mini | OpenAI | Model | 16 Apr 2025 | 39.82 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Qwen 3.5 Flash | ALAlibaba | Model | Not stated | 39.82 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| MiMo V2.5 | XIXiaomi | Model | Not stated | 39.62 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Mistral Medium 3.5 | Mistral | Model | 28 Apr 2026 | 39.62 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Qwen 3 Max Thinking | ALAlibaba | Model | Not stated | 39.52 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| GLM 4.7 | Zhipu | Model | 22 Dec 2025 | 39.32 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Grok 4.1 Fast (Reasoning) | SpaceX AI | Model | 19 Nov 2025 | 39.32 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| GPT-5 nano | OpenAI | Model | 7 Aug 2025 | 39.22 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Grok 4.1 Fast Non-Reasoning | SpaceX AI | Model | 19 Nov 2025 | 39.22 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Ling 3.0 Flash Fin | AGAnt Group | Model | Not stated | 39.22 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| MiMo V2.5 Pro | XIXiaomi | Model | Not stated | 38.63 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), PhysicianBench |
| MAI-Thinking-1 | Microsoft | Model | 12 Aug 2026 | 38.51 of 12 boards | 1 / 14 | HealthBench Professional |
| Gemini 2.5 Flash Lite | Google | Model | 22 Jul 2025 | 38.42 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Llama 4 Maverick | Meta | Model | 5 Apr 2025 | 38.12 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Qwen3.6 Plus | ALAlibaba | Model | 1 Apr 2026 | 37.94 of 12 boards | 4 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM), PhysicianBench |
| Gemma 4 12B | Google | Model | 3 Jun 2026 | 37.21 of 12 boards | 1 / 14 | MedXpertQA (MM) |
| Qwen3-VL-235B-A22B | ALAlibaba | Model | Not stated | 36.81 of 12 boards | 1 / 14 | MedXpertQA (MM) |
| Mercury 2.5 | INInception | Model | 8 Sep 2026 | 36.32 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| MiniMax-M2.7 | MiniMax | Model | 18 Mar 2026 | 35.33 of 12 boards | 3 / 14 | MedScribe (Vals AI), MedCode (Vals AI), PhysicianBench |
| Laguna M.1 | POPoolside | Model | 28 Apr 2026 | 35.12 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Laguna XS.2 | POPoolside | Model | 28 Apr 2026 | 33.62 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Grok 4.20 | SpaceX AI | Model | 10 Mar 2026 | 32.64 of 12 boards | 4 / 14 | MedScribe (Vals AI), MedCode (Vals AI), MedXpertQA (MM), PhysicianBench |
| Llama 4 Scout | Meta | Model | 5 Apr 2025 | 32.42 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Command A+ | COCohere | Model | 20 May 2026 | 31.92 of 12 boards | 2 / 14 | MedScribe (Vals AI), MedCode (Vals AI) |
| Claude Haiku 4.5 | Anthropic | Model | 15 Oct 2025 | 31.05 of 12 boards | 6 / 14 | HealthBench Professional, MedScribe (Vals AI), MedCode (Vals AI), PhysicianBench, CHI-Bench, RCM-Bench |
| Gemma 4 E4B | Google | Model | 2 Apr 2026 | 29.21 of 12 boards | 1 / 14 | MedXpertQA (MM) |
| Gemma 4 E2B | Google | Model | 2 Apr 2026 | 27.11 of 12 boards | 1 / 14 | MedXpertQA (MM) |
| Nemotron 3.5 Lightning | NVIDIA | Model | 11 Aug 2026 | 14.71 of 12 boards | 1 / 14 | MedScribe (Vals AI) |
| GPT-5.4 mini | OpenAI | Model | 17 Mar 2026 | Not ranked | 2 / 14 | HealthAgentBench, CHI-Bench |
| Doximity Ask 6.1 | DODoximity | Product | Not stated | Not ranked | 1 / 14 | First, Do NOHARM (v2) |
| Glass 5.6 Max | GHGlass Health | Product | Not stated | Not ranked | 1 / 14 | First, Do NOHARM (v2) |
| GPT-5.3-Codex | OpenAI | Model | 5 Feb 2026 | Not ranked | 1 / 14 | HealthAgentBench |
| LiSA 2.5 | AMAMBOSS | Product | Not stated | Not ranked | 1 / 14 | First, Do NOHARM (v2) |
| MedGuard | MEMedGuard | Product | Not stated | Not ranked | 1 / 14 | CHI-Bench |
| NVIDIA Nemotron 3 Ultra 550B A55B | NVIDIA | Model | 4 Jun 2026 | Not ranked | 1 / 14 | CHI-Bench |
| OpenEvidence | OPOpenEvidence | Product | Not stated | Not ranked | 1 / 14 | First, Do NOHARM (v2) |
| Qwen3.6-Max | ALAlibaba | Model | Not stated | Not ranked | 1 / 14 | CHI-Bench |