Hey All,
We just wrapped a hands-on round with our FinanceEval framework: here’s what I discussed in the video and my current top open-source picks for finance-advice–focused models on Hugging Face:
Top Open Source Finance Models – BrainDrive
Top 3 (with quick stats)
meta-llama/llama-3-70b-instruct · Hugging Face
Score 6.26 | 6-metric profile (Trust, Accuracy, Explainability, Client-First, Risk Safety, Clarity) | 70B params | Meta license | EN
https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct
meta-llama/llama-3.3-70b-instruct · Hugging Face
Score 5.87 | 6-metric profile | 70B params | Meta license | EN
https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct
nvidia/llama-3.1-nemotron-70b-instruct · Hugging Face
Score 5.78 | 6-metric profile | 70B params | NVIDIA Open Model License | EN
https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct
How we ranked them (FinanceEval by BrainDrive)
FinanceEval is our evaluation workflow for AI-generated financial advice.
We score models on 6 practical metrics—from Trust & Transparency, Competence & Accuracy, Explainability, and Client-Centeredness to Risk Safety and Clarity & Financial Literacy Support.
Individual scores roll up into a weighted total, which determines ranking.
Docs (scoring & math):
https://github.com/BrainDriveAI/ModelMatch/tree/main/FinanceEval/Docs
Workflow we used
Model shortlist (20+ HF candidates) → Multi-domain prompts (budgeting, investing, taxation, retirement) → Responses per model → FinanceEval scoring → Weighted aggregation → Ranking.
Try it yourself
Code toolkit: https://github.com/BrainDriveAI/ModelMatch/tree/main/FinanceEval
No-code evaluator: https://huggingface.co/spaces/BrainDrive/FinanceEval
About ModelMatch
ModelMatch helps you discover the most suitable open-source model for your domain and task—starting with summarization, expanding into therapy, email generation, and now finance evaluation.
If you test other models or get different results, ping us; happy to compare notes.
Regards,
Navaneeth