The all-in-one financial operations platform that saves businesses time and money. Trusted by 30,000+ teams.
Mar 11 • 5 tweets • 2 min read
Word on the timeline is that agents will go from automating coding to knowledge work in 2026. So we benchmarked frontier LLMs on doing financial tasks to see what's good.
The result: American models are outperforming their Chinese counterparts in both reliability + performance.
Gemini is the king of multimodal, with near-identical performance on contextual OCR to GPT-5.1-high, at less than 1/3 the cost.
