👩🍳 How we use AI at Tech in Asia, thoughtfully and responsibly.
🧔♂️ A friendly human may check it before it goes live. More news here
🧔♂️ A friendly human may check it before it goes live. More news here
Mistral AI revenue climbs to over $400m
French AI startup Mistral reports its revenues have increased 20x over the past year, reaching an annualized rate of over US$400 million, up from US$20 million a year earlier.
The company, founded in 2023 and based in Paris, aims to surpass US$1 billion in annual revenue by year-end by expanding its enterprise customer base to over 100 firms.
Mistral, which raised €1.7 billion (US$2 billion) in September led by ASML, plans to invest €1.2 billion (US$1.4 billion) to build AI data centers in Sweden, its first outside France.
The facility will offer 23 MW of computing power and is expected to generate more than €2 billion (US$2.3 billion) in revenue over five years.
🔗 Source: Financial Times
🧠 Food for thought
Implications, context, and why it matters.
Mistral’s revenue surge contrasts with a crowded field of top-performing AI models
- Mistral’s revenue is climbing fast, yet its models face a long list of strong rivals.
- As of December 2025, LMArena (a community-run benchmark that ranks AI models based on human-preference votes) ranks several models near the top 1. Google’s Gemini 3 Pro sits alongside xAI’s Grok 4.1; Anthropic’s Claude Opus 4.5; OpenAI’s GPT-5.2.
- That tight race rewards models that win in a clear lane. Anthropic’s Claude Opus 4.5 is the first model to clear 80% on the SWE-bench Verified coding test (a benchmark that measures how well AI models solve real software engineering tasks) 1.
The push for sovereign AI extends far beyond Europe
- Mistral is building independent European AI infrastructure as part of a broader push for sovereign AI that cuts reliance on foreign cloud and AI providers.
- In the Middle East, Khazna Data Centers (a United Arab Emirates-based data center operator) said its Ajman facility (QAJ1) is built to become the region’s first AI-optimized data centre 2.
- The effort also reaches hardware. MLQ.ai reports that China’s Zhipu AI has trained an image generation model only on Huawei chips to reduce reliance on foreign hardware 3.




