Ant Group opens a 124B finance MoE and a search-agent benchmark
- Source
- AntLingAGI
- Date
We’re open-sourcing Ling-3.0-flash-Fin, a finance-enhanced model for real-world workflows, and FinFIRST, an expert-built benchmark for financial search agents. Two open releases, one goal: making financial AI more accessible and verifiable.
Ling-3.0-flash-Fin is a 124B-parameter MoE with 5.1B active parameters and a 256K context window. With open weights, teams can deploy it privately, connect it to search, Python, databases and spreadsheets, and adapt it to their own financial workflows. Hugging Face: https://t.co/6d8cgjAdm9 ModelScope: https://t.co/6Us0krfcaU
FinFIRST is an open benchmark for financial search agents, built by Ant Group with professional support from the investment banking team at China International Capital Corporation Limited (CICC). V1 includes 123 expert-authored tasks, 701 atomic criteria and 12,300 rubric points for tracing evidence, sources and calculations. Hugging Face Dataset: https://t.co/aSKFvquYUS ModelScope Dataset: https://t.co/GuUE1wmKC8 Technical Report:
Join the AntLing community on Discord for release updates, deployment discussions, feedback and collaboration: https://t.co/TwbjGjpDza We’d love to see what you build with Ling-3.0-flash-Fin and FinFIRST.
- mixture-of-experts — A model built from many specialist sub-networks where only a few activate per token, giving big-model capability at small-model running cost.
- context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
- AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
- open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.
let teams run a finance-tuned privately and connect it to search, Python and databases, while FinFIRST supplies rubric-scored tasks for measuring whether a financial search actually retrieves and reasons correctly.
Checking sign-in…
Loading comments…