---
format: "aidr-story-markdown/v1"
id: "2981768d546156bf30cd9956e2f931a02e53514fd05e60d2b15c6dfd488d28aa"
canonical_url: "https://aidr.today/2981768d?lang=en"
title: "NEAR AI Tops Lean Eval v1 After 4 Weeks of Competition"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-02T21:39:00.000Z"
category: "Tools"
topics: ["deepseek","agent","devtools"]
source_urls: ["https://marketbrief.now/ai/near-ai-tops-lean-eval-v1-after-4-weeks-of-competition-6ffd214a","https://huggingnews.com/ai/near-ai-tops-lean-eval-v1-after-4-weeks-of-competition-6ffd214a"]
summary: "An AI agent developed by NEAR's research team, powered by the open-weight DeepSeek V4.1 Flash model, topped the Lean Eval v1 benchmark for difficult formalization problems four weeks after joining the competition."
---

# NEAR AI Tops Lean Eval v1 After 4 Weeks of Competition

> [Open the canonical story](<https://aidr.today/2981768d?lang=en>)

**Published:** 2026-10-02T21:39:00.000Z
**Category:** Tools
**Topics:** deepseek, agent, devtools

## Summary

An AI agent developed by NEAR's research team, powered by the open\-weight DeepSeek V4\.1 Flash model, topped the Lean Eval v1 benchmark for difficult formalization problems four weeks after joining the competition\.

## Sources

- [Story source](<https://marketbrief.now/ai/near-ai-tops-lean-eval-v1-after-4-weeks-of-competition-6ffd214a>)
- [Story source](<https://huggingnews.com/ai/near-ai-tops-lean-eval-v1-after-4-weeks-of-competition-6ffd214a>)

