Flower AI Summit 2026·April 15-16·London
Medical

Medical LLM Leaderboard

Embrace federated LLM fine-tuning on medical tasks and secure your spot on the leaderboard!

In the domain of Medical AI, developing models that can accurately interpret clinical notes, answer medical questions, and support healthcare professionals is of paramount importance.

Federated LLM fine-tuning on medical tasks enables hospitals and research institutions to collaborate without sharing sensitive patient data.

This approach ensures fine-tuned models are robust across diverse medical benchmarks and can generalize to real-world clinical scenarios.

RankTeamCodeBase ModelComm. CostsAverage (↑)PubMedQAMedMCQAMedQACareQADate
1ZeroOne.AILlama3.1-Aloe-Beta-8B2.0 GB63.5774.8055.3959.3164.7914.01.26
2FL-finetune-JB-DCBio-Medical-Llama-3-8B1.6 GB63.1270.4060.9365.8255.3612.04.25
3AI4EOSC TeamBio-Medical-Llama-3-8B1.0 GB62.1466.2060.2968.4253.6425.04.25
4Gachon Cognitive Computing LabBio-Medical-Llama-3-8B2.0 GB62.1265.8060.3868.5753.7619.03.25
5mHealth LabBio-Medical-Llama-3-8B0.7 GB62.0365.8068.3460.3153.6721.04.25
6ZJUDAILlama-3.1-8B-Instruct2.0 GB61.7559.9469.4059.9457.7404.04.25
7mHealth LabBio-Medical-Llama-3-8B2.9 GB61.4669.6059.1263.9453.1821.04.25
8AI4EOSC TeamBio-Medical-Llama-3-8B2.0 GB61.1470.8058.0462.8452.8706.03.25
9FL-finetune-JB-DCBio-Medical-Llama-3-8B1.6 GB60.7570.2059.2959.2354.3112.04.25
10ZJUDAIQwen2.5-7B-Instruct1.5 GB60.6465.1556.8065.1555.4604.04.25
11Gachon Cognitive Computing LabBio-Medical-Llama-3-8B2.0 GB60.3770.6057.6861.5051.6821.11.24
12ZJUDAIMistral-7B-Instruct-v0.32.0 GB53.4554.4055.2054.4049.8004.04.25
13Massimo R. ScamarciaQwen2.5-7B-Instruct45.1 GB49.0744.6039.0655.4657.1623.11.24
14Abdelkareem ElkhatebQwen2.5-1.5B-Instruct11.0 GB46.0052.6040.8140.2250.3802.03.25
15Massimo R. ScamarciaMistral-7B-Instruct-v0.345.1 GB40.9563.8025.6039.8334.5723.11.24
16mHealth LabJSL-MedLlama-3-8B-v2.00.3 GB39.3058.6027.5030.3040.8021.04.25
17mHealth LabLlama-medx_v3.218.6 GB39.0439.4036.3441.5638.8721.04.25
18BaselineMistral-7B-v0.340.7 GB32.8259.0023.6927.1021.4701.10.24
19mHealth Labmeerkat-7b-v1.018.5 GB22.3036.0013.6725.7713.7621.04.25
20Massimo R. ScamarciaQwen3-0.6B8.1 GB21.9420.4020.1320.3526.8604.05.25
21T-IoI@UNRSmolLM2-135M-Instruct1.4 GB17.0254.206.930.096.8614.04.25
22T-IoI@UNRSmolLM2-135M1.4 GB11.297.2016.802.6718.5114.04.25
23T-IoI@UNRSmolLM2-360M-Instruct2.5 GB9.3722.007.141.337.0414.04.25