US EditionMistral's latest open-source release says smaller models beat large ones - here's why
16 Articles
16 Articles
The company said it trained Mistral Large 3 in many languages, making advanced AI usable for billions of people with different native languages.
While the Silicon Valley giants are fighting with billions of dollars and big servers like cities, our French nugget Mistral comes out of its hat the Mistral family 3. The goal? Prove that we can play in the courtyard of large "border" models without however mortgageing its infrastructure over ten generations. On the menu: a wide multimodal model for the elite, and an armada of small nervous models for everyday life.
Mistral Unveils Its Most Powerful Open Model Yet with Mistral 3
Mistral AI has launched Mistral 3, a new generation of open multimodal and multilingual models that includes three compact “Ministral” systems and its most advanced release yet, Mistral Large 3. All models are available under the Apache 2.0 licence. The flagship Mistral Large 3 is a sparse mixture-of-experts model trained on 3,000 NVIDIA H200 GPUs, featuring 41 billion active parameters and 675 billion total parameters. The company claims that …
NVIDIA and Mistral AI Bring 10x Faster Inference for the Mistral 3 Family on GB200 NVL72 GPU Systems
NVIDIA announced today a significant expansion of its strategic collaboration with Mistral AI. This partnership coincides with the release of the new Mistral 3 frontier open model family, marking a pivotal moment where hardware acceleration and open-source model architecture have converged to redefine performance benchmarks. This collaboration is a massive leap in inference speed: the new models now run up to 10x faster on NVIDIA GB200 NVL72 sys…
Mistral Releases New Family of AI Models, Promoting Performance-to-Cost Ratio
Artificial intelligence (AI) startup Mistral released the next generation of its models Tuesday (Dec. 2), saying they “represent the best performance-to-cost ratio in their category.” The Mistral 3 models include three small, dense Ministral models (14B, 8B and 3B) and the company’s most capable model to date, Mistral Large 3, the company said in a Tuesday press release. Mistral Large 3 is trained from scratch on 3,000 Nvidia H200 GPUs…
Coverage Details
Bias Distribution
- 100% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
