Thomson Reuters unveils its first in-house LLM Thomson-1.0

Thomson Reuters has launched its first in-house LLM, Thomson-1.0, at ILTACON2026, fine-tuned for legal and tax on 175 years of proprietary data and claimed to rival frontier models. Built on a realigned version of Alibaba's open-source Qwen, the Large model cost only $450,000 to train.

2026-08-24 ~ 2026-08-26 · 4 related posts

• Thomson Reuters Launches Proprietary LLM 'Thomson' at Fraction of Frontier Costs — schwarzjn_ · 2026-08-24
• Thomson Reuters Launches Thomson-1.0 Legal LLM — josh_wills · 2026-08-25
• Thomson Reuters builds Thomson-1 model on Alibaba's Qwen to cut reliance on Anthropic — kuanhoong · 2026-08-26
• Thomson Reuters releases Thomson-1.0-Small for law and tax — RedditUsr2 · 2026-08-26