DeepSeek releases first AI capable of top score at maths Olympiad

(Web Desk) - Chinese startup DeepSeek has released the first ever open AI model capable of achieving a gold medal at the prestigious International Mathematical Olympiad.

The Math-V2 model reached the gold-level score at this year’s IMO – a feat that only 8 per cent of human participants achieved – by demonstrating its reasoning abilities rather than just produce a simple answer.

DeepSeek has now made the model available to the public on the developer platforms Hugging Face and GitHub, allowing anyone to run or modify it for free.

“Imagine owning the brain of one of the best mathematicians in the world for free to explore it for research, fine-tune it, optimise it, run it on your own hardware,” said Hugging Face co-founder and chief executive Clement Delangue.

“No limitations, no nerfing [reducing the power], no company or government to take it back. That’s democratisation of AI and knowledge at its best, literally.”

Models built by Google DeepMind and ChatGPT creator OpenAI have both scored gold-level performances at the annual IMO, but neither company has made them open-source.

DeepMind’s model is currently available for subscribers to its premium Ultra package, while OpenAI is yet to publicly release its model.

DeepSeek’s model proved its ability to “self-verify”, according to the firm’s researchers, allowing it to solve mathematical problems without known solutions.

This addresses a key bottleneck in current AI systems, which tend to improve only on tasks with easily verifiable solutions, the researchers said.

DeepSeek launched earlier this year, making headlines for claims that its large language model (LLM) was created at a fraction of the cost of rivals like OpenAI’s ChatGPT.

It quickly becoming the most downloaded free app in the US, leading President Donald Trump to describe it as a “wake-up call” for the tech industry.