TBC
FRIDAY 10 JULY 2026
Paris Open Source
AI Summit is back
AI Summit is back
Registration is now open
Become an AI Summit participant
Become an AI Summit participant
Get in early to book your seats at the biggest AI summit in Paris.
Days
Hours
Minutes
Seconds
About POSAIS
There are conferences where people talk about AI. Then there's the day where the people who actually build open source AI show up, stand up, and speak up. This is what POSAIS is.
POSAIS is a dedicated day for Open Source AI “doer”. This summit is for made for CTO’s, CIO’s, scientists and academics to meet, discuss and challenge themself.
POSAIS is a dedicated day for Open Source AI “doer”. This summit is for made for CTO’s, CIO’s, scientists and academics to meet, discuss and challenge themself.
Watch Last Year Video :
Exceptional location
Maison de la Radio et de la Musique
116 Av. du Président Kennedy, 75016 Paris
116 Av. du Président Kennedy, 75016 Paris
Program
Five streams. Zero filler.
- Geopolitics & sovereignty :Why mastering open source AI is a strategic imperative, not a tech preference. The real power map and where Europe stands.
- State of the artLive demos, honest benchmarks, hard conversations about what actually works at scale. Models, infrastructure, datasets, deployment. Hands in the code.
- Building the coalitionHow we federate, fund, and sustain open source AI as a global commons. The economic models that work. The standards we need. The alliances that matter.
- Impact & empowermentHow open source AI gives power back. To the Global South. To education. To communities that were never supposed to have a seat at the table.
- New frontiers: Physical AIAI leaves the cloud and enters the real world. We open with a live demo of a Unitree humanoid robot on stage, then dive into embodied AI, open source VLA models, sim-to-real transfer, and what happens when the next wave of AI has legs, arms, and hands. This is where open source meets robotics and it's happening faster than anyone expected.
Speakers
Coming soon
REPLAY 2025 : Take a look at last yeat keynotes & Round-tables
Join global leaders, innovators, and policymakers as we lay the groundwork for shaping the future of AI.
Be part of history! Watch live as if you were there
The First Official Global Meetup of the OpenLLM Europe Community
Dive into the heart of innovation with two exclusive parallel events.
Share Your Vision at the Paris Open Source AI Summit
We invite experts to propose talks on Open Source AI, ethics, public policy, and global impact projects aligned with the event’s goals.
Register now to participate
Just type your contacts
We will contact you shortly with more details
Become a Summit Partner
Just type your contacts
We will contact you shortly with more details
Cookie management
We respect your privacy and aim for the best website experience in compliance with GDPR. Allowing cookies enables a tailored experience. For more information, please read our Privacy Policy.
We respect your privacy and aim for the best website experience in compliance with GDPR. Allowing cookies enables a tailored experience. For more information, please read our Privacy Policy.
Hackathon - Instruction Dataset for LUCIE
Making a universal IA model better by "Instruction dataset for LUCIE" hackathon: a collaborative initiative for a truly LLM aligned with european and ethical values
This sequence is exclusively dedicated to the Open Source Large Language Model, LUCIE. Throughout the day, participants will have the opportunity to explore the journey of LUCIE and participate to the hackathon. This hackathon aims to mobilize the OpenLLM France community and other participants to create the largest dataset of instructional data in French, designed to train Open Source models like LUCIE. The goal is to collect, annotate, and create a set of 100,000 to 200,000 Q/A pairs.
This sequence is exclusively dedicated to the Open Source Large Language Model, LUCIE. Throughout the day, participants will have the opportunity to explore the journey of LUCIE and participate to the hackathon. This hackathon aims to mobilize the OpenLLM France community and other participants to create the largest dataset of instructional data in French, designed to train Open Source models like LUCIE. The goal is to collect, annotate, and create a set of 100,000 to 200,000 Q/A pairs.