Home · World · Oct 6 archive
Mistral CEO says new AI model beats Chinese ones in some areas
Confirmed
In Short: At an Abu Dhabi conference, Mistral AI's CEO said the company's latest model outperforms Chinese competitors in cybersecurity, challenging the narrative that Europe cannot compete.
At an Abu Dhabi conference on October 6, Mistral AI's co-founder and CEO, Arthur Mensch, claimed that the company's newest AI model outperforms Chinese rivals in cybersecurity.
Mensch stated, 'The model we're announcing today is above the Chinese models on certain aspects, including cyber. So the narrative that Europe cannot compete is something that is not true.'
Mensch's comments came days after Mistral AI secured a €3 billion funding round, raising its valuation to €21 billion.
The company, founded just three years ago, has positioned itself as a competitor to both US and Chinese AI models, offering tools and services that businesses can use to run AI.
Mensch emphasized that Mistral's position outside the US and China could help it win customers in regions such as the Gulf and Asia-Pacific, where buyers are wary of choosing between American and Chinese infrastructure.
However, Mensch did not provide specific model names, benchmarks, or scores to substantiate his claims, leaving the verification of his statements to be determined by real-world customer usage.
Mistral AI's business plan extends beyond launching new models; it also includes providing computing power and building tools around customers' own data.
The company's latest model, Mistral Large 3, released in December 2025, showed promise by outperforming DeepSeek V3.1 on several metrics, though it trailed the newer V3.2 on the LMArena leaderboard.
Despite the claims, the real test will come when customers start using the model and its results can be compared in real-world work.
Mensch's comments come at a time when the narrative of China's dominance in open models has gained momentum, making his statements a direct challenge to that narrative.
Mensch added, 'There's an enormous desire and an enormous need for alternative technology suppliers,' pitching the company directly at buyers in the Gulf and Asia-Pacific who do not want to choose between American and Chinese infrastructure.
What's confirmed
- At an Abu Dhabi conference on October 6, Mistral AI's co-founder and CEO, Arthur Mensch, claimed that the company's newest AI model outperforms Chinese rivals in cybersecurity.
- Mensch's comments came days after Mistral AI secured a €3 billion funding round, raising its valuation to €21 billion.
- The company, founded just three years ago, has positioned itself as a competitor to both US and Chinese AI models, offering tools and services that businesses can use to run AI.
- Mensch emphasized that Mistral's position outside the US and China could help it win customers in regions such as the Gulf and Asia-Pacific, where buyers are wary of choosing between American and Chinese infrastructure.
- However, Mensch did not provide specific model names, benchmarks, or scores to substantiate his claims, leaving the verification of his statements to be determined by real-world customer usage.
- Mistral AI's business plan extends beyond launching new models; it also includes providing computing power and building tools around customers' own data.
- The company's latest model, Mistral Large 3, released in December 2025, showed promise by outperforming DeepSeek V3.1 on several metrics, though it trailed the newer V3.2 on the LMArena leaderboard.
- Despite the claims, the real test will come when customers start using the model and its results can be compared in real-world work.
- Mensch's comments come at a time when the narrative of China's dominance in open models has gained momentum, making his statements a direct challenge to that narrative.
- Mensch added, 'There's an enormous desire and an enormous need for alternative technology suppliers,' pitching the company directly at buyers in the Gulf and Asia-Pacific who do not want to choose between American and Chinese infrastructure.
What's still developing
- However, the company often faces challenges from the flagship US rivals like OpenAI and Anthropic, and even from the low-cost Chinese models.
- Chinese AI companies have made a name for themselves by offering capable models at lower costs.
- Mensch's comment reads less like a stray boast and more like a direct shot across that bow, timed to a moment when the "China is winning open models" narrative has real momentum.
- Mistral closed a €3 billion funding round in September at a €21 billion valuation, led by Samsung Electronics with EQT's Scaleup Europe Fund and PSG Equity co-leading.
- The deal raises fresh questions about what European AI sovereignty actually means when the compute and capital still come from... - Microsoft GPU infrastructure investment Europe - Mistral AI cloud services partnership Here's the thing: without a named model, a named benchmark, and a published score, there's nothing here a reader, or a competitor, can actually verify.
- StartupFortune has covered three separate Chinese open-model stories in recent weeks alone, from GLM-5.3 landing inside Amazon Bedrock to Reflection AI's Beam positioning itself as an answer to China's compute-efficient approach.