Study finds Manus model automates 2.5% of $10,000 freelance jobs
A new study found that AI models are still pretty far from replacing freelancers.
When tested on complex jobs like game development and architectural design, the kind that cost $10,000 and took them more than 100 hours to complete, Manus, the best-performing model, achieved an automation rate of 2.5%.
Models struggle with learning and vision
Among the models tested, Manus led the pack but completed just 2.5% of tasks. Others like Grok 4, Sonnet 4.5, GPT-5, ChatGPT agent, and Gemini 2.5 Pro scored even lower. Gemini finished less than 1%.
The researchers pointed out that current AIs struggle especially with learning new things and handling visual tasks.
Still, they noted progress is happening: "While absolute automation rates are low, our analysis shows that models are steadily improving and that progress on these complex tasks is measurable," the researchers said.