Pleias

Paris lab building small open LLMs and the largest rights-cleared training dataset in Europe

Quick Overview

Company Pleias SAS
Category LLMs & AI
Headquarters Paris, France
Founded 2023
EU Presence EU (France)
Open Source Apache 2.0 (Pleias small language models)
Pricing Common Corpus dataset and Pleias small language models (Pleias-RAG-1B, Pleias-Pico and others) free and open on Hugging Face under Apache 2.0; Synth and Stratum on-premise deployments priced through a demo call, no published rate card.
Free Option Free open models and Common Corpus dataset on Hugging Face; book a demo for Synth/Stratum
Replaces Proprietary LLM training-data vendors, closed synthetic-data tools

Detailed Review

Pros and Cons

Pros

  • Common Corpus: largest open, rights-cleared LLM training dataset
  • Small open-weight models on Hugging Face under Apache 2.0
  • Production track record (RATP's Sillon model)
  • On-premise deployment for Synth and Stratum
  • Backed by Nvidia, Mozilla and the Wikimedia Foundation as partners

Cons

  • No general-purpose flagship chat product
  • No public pricing for Synth or Stratum
  • Small, young company founded in 2023
  • Best fit is data and training pipelines, not an end-user assistant

Alternatives to Pleias

See all LLMs & AI →

Frequently Asked Questions

What is Pleias?

Pleias occupies a different corner of the European AI map: instead of chasing frontier scale, it builds the data layer underneath — and increasingly, the small models trained on it.

Common Corpus, its flagship dataset, is described as the largest open, rights-cleared collection for LLM pre-training, built from government records, legal archives and scientific literature with clear provenance rather than scraped copyrighted text. The work won an ICLR Oral, and partners include Nvidia, Mozilla, the Wikimedia Foundation and the AI Alliance.

Where is Pleias based?

Pleias operates from Paris, France, which places it under EU (France).

What does Pleias cost?

Common Corpus dataset and Pleias small language models (Pleias-RAG-1B, Pleias-Pico and others) free and open on Hugging Face under Apache 2.0; Synth and Stratum on-premise deployments priced through a demo call, no published rate card.. Free open models and Common Corpus dataset on Hugging Face; book a demo for Synth/Stratum.

Who is Pleias best for?

Teams needing open, rights-cleared training data and small models. Training data with known, rights-cleared provenance — the opposite of "trust us" that most LLM pre-training asks for.

What are the drawbacks of Pleias?

No general-purpose flagship chat product. No public pricing for Synth or Stratum. Small, young company founded in 2023. Best fit is data and training pipelines, not an end-user assistant.

Who worked on this review

Three people touch every tool page: one writes it, a second edits it, and a third checks the compliance and pricing claims against the vendor's own documentation.

Ingrid Halvorsen
Written by

Ingrid Halvorsen

Managing Editor · Oslo, Norway

Runs the review process and decides when a page is ready to publish or needs another pass.

Daniel Brandt
Edited by

Daniel Brandt

Privacy & Compliance Researcher · Berlin, Germany

Checks the compliance claims: where the company is established, where the data sits, and what the DPA actually says.

Sebastiaan Smits
Fact-checked by

Sebastiaan Smits

Founder & Editor · Netherlands

Selects the tools, writes the reviews, and checks where each company is actually established.

Read our editorial process for how we source, verify and update these pages — and how we keep affiliate income separate from what we recommend.

Go to Pleias