Post-Cutoff

Arthur Mensch: “We’re building AGI from this train station”

Arthur Mensch @arthurmenschX

Why it matters

Mistral CEO’s joke about where the Mistral Large 4 training cluster is (~130k views by 7 Oct).

Summary

Posted 17:56 UTC on 6 Oct, quoting Guillaume Lample’s 2/n tweet (~649k views by 7 Oct): ML4 “was trained on 3,800 NVIDIA Grace Blackwell GPUs in our European datacenters -- our cluster in Bruyères-le-Châtel built using our Series B fundraise”, with Series C and D clusters “coming online soon”. Lample’s tweet shows a photo of the Breuillet – Bruyères-le-Châtel railway station (the station sign is legible), and Mensch’s line refers to that photo. Reach at 11:31 UTC on 7 Oct: ~130k views, 3,416 likes.

Archived text

We’re building AGI from this train station

Quoting @GuillaumeLample: ML4 was trained on 3,800 NVIDIA Grace Blackwell GPUs in our European datacenters -- our cluster in Bruyères-le-Châtel built using our Series B fundraise. We are investing heavily in infrastructure, with our Series C and D clusters coming online soon. This additional compute will let us train longer, pursue more ambitious post-training, and iterate faster. We expect large and rapid improvements in the weeks and months ahead. 2/n

Source: x.com/arthurmensch/status/2107530509366538373Archived 2026-10-07 via manual (api.fxtwitter.com, fetched 11:31 UTC).Counts: 129,666 views, 3,416 likes, 222 reposts, 114 replies, 32 quotes (at fetch time)

Cited in

  1. Model releases 98 days after the cutoff

    Mistral releases Mistral Large 4 (“le Chonk”)

People in this post

Guillaume Lample, Co-founder and Chief Scientist, Mistral AI