GLM 5.2 is nearly as accurate as a human book keeper

·Hacker News··

We benchmarked GLM 5.2, an open-weights AI model, on a real quarterly UK VAT return: 59 transactions, 68 minutes, $2.73 of tokens — and a net VAT position within 7 pence of the human-prepared return.

Read full article →

Related Articles

Coconut oil jet fuel matches kerosene's efficiency in engine tests
mdp2021 · Hacker News · 8h ago
Slovakia finds Russian backdoor in traffic speed cameras
dredmorbius · Hacker News · 9h ago
GLM-5.3 (open-weight) beat Anthropic/OpenAI models – for 1/5 the cost
ed-is-ai · Hacker News · 7h ago
There's no reason for software to be slow anymore
Jach · Hacker News · 1d ago
How Complex Systems Fail (1998)
shortcrct · Hacker News · 9h ago