Writings
partnerships
How Zapier Used Lab to Build AutomationBench
April 20, 2026
partnerships
Launching FrontierSWE on the Environments Hub
April 16, 2026
partnerships
Partnering with Browserbase to Train Browser and Computer Use Agents
March 30, 2026
partnerships
Leveraging NVIDIA to Build the Open Superintelligence Stack
March 16, 2026
announcements
Introducing Lab: The Full-Stack Platform for Training your Own Models
February 10, 2026
.png&w=3840&q=75)
research
Recursive Language Models: the paradigm of 2026
January 1, 2026
announcements
INTELLECT-3: A 100B+ MoE trained with large-scale RL
November 26, 2025
announcements
Scaling Our Open-Source Environments Program
October 27, 2025
announcements
Environments Hub: A Community Hub To Scale RL To Open AGI
August 27, 2025
announcements
SYNTHETIC-2 Release: Four Million Collaboratively Generated Reasoning Traces
July 10, 2025
announcements
SYNTHETIC-2
June 23, 2025
announcements
PCCL - Prime Collective Communications Library
May 20, 2025
announcements
INTELLECT-2 Release: The First 32B Parameter Model Trained Through Globally Distributed Reinforcement Learning
May 11, 2025
announcements
Planetary-Scale Inference: Building a Distributed Inference Engine for the Public Internet
April 28, 2025
.png&w=3840&q=75)
announcements
INTELLECT-2: The First Globally Distributed Reinforcement Learning Training of a 32B Parameter Model
April 15, 2025
announcements
$15M to Build The Open Superintelligence Stack
February 28, 2025
-min.png&w=3840&q=75)
announcements
SYNTHETIC-1 Release: Two Million Collaboratively Generated Reasoning Traces from Deepseek-R1
February 20, 2025
announcements
SYNTHETIC-1: Scaling Distributed Synthetic Data Generation for Verified Reasoning
February 6, 2025
announcements
TOPLOC: A Locality Sensitive Hashing Scheme for Trustless Verifiable Inference
January 28, 2025
announcements
Distributed Training in the Inference-Time-Compute Paradigm
January 21, 2025
announcements
METAGENE-1: Metagenomic Foundation Model
January 6, 2025
announcements
INTELLECT-1 Release The First Globally Trained 10B Parameter Model
November 29, 2024
%2520(1)-1.png&w=3840&q=75)
announcements
INTELLECT–1: Launching the First Globally-Distributed Training of a 10B Parameter Model
October 11, 2024
announcements
OpenDiLoCo: An Open-Source Framework for Globally Distributed Low-Communication Training
July 11, 2024
announcements
Introducing Prime Intellect Compute: The compute exchange
July 1, 2024
.png&w=3840&q=75)
announcements
State-of-the-Art in Globally-Distributed Training
April 23, 2024