[ Tech stack ]
Meta Llama
Meta's open-weight LLMs, multilingual and hostable wherever you want.
Meta Llama is Meta's family of open-weight LLMs: Llama 3.x (1B to 405B) and Llama 4 (multimodal MoEs) span edge-sized to very large models, multilingual and under a community license that allows commercial use. The weights can be served locally with vLLM or Ollama, or called through the cloud providers that host them. It's our choice when an AI project needs to keep its data and model under control, or fine-tune an LLM on its own domain.
[ Why Meta Llama at Dexon ]
What this technology does well, and why we use it.
Typical usage: Self-hosted LLM for a RAG pipeline or assistant, domain fine-tuning, cost-controlled inference.
- 01
Open weights and community license: commercial use and fine-tuning allowed.
- 02
Llama 3.x and Llama 4 in several sizes: from edge models (1B, 3B) to large MoEs.
- 03
Multilingual, with a very active ecosystem of tools and variants.
- 04
Self-host with vLLM or Ollama, or call it through cloud providers.
[ Complementary technologies ]
The building blocks we often mobilise alongside.
A stack rarely exists alone. Here are the technologies Dexon most often pairs with this one, through pipeline habits, usage similarity or internal mastery. Click on a building block to see its scope.
[ Reassurance ]
- 0+
- custom projects delivered
- 30
- engineers, designers, project managers
- 80 %
- from top French schools
- 24 h
- average reply time
[ They trust us ]
More than 100 French and European companies trust us
[ Press ]
They talk about us.
Nationwide coverage by BFM Business, Le Figaro, Challenges, La Tribune and CNews. An outside reading of our work and our innovations.
