Product
Prime Intellect announced on 1 July 2024 that Prime Intellect Compute, its platform for aggregating and orchestrating global GPU resources, was fully public.
Supported
Project announcement
Evidence (1): 1 supporting
- supports: Prime Intellect
Prime Intellect Compute, our platform for aggregating and orchestrating global GPU resources, is now fully public.
Prime Intellect announced Lab on 10 February 2026, a full-stack platform for agentic post-training that unifies its Environments Hub with hosted training and hosted evaluations.
Supported
Project announcement
Evidence (1): 1 supporting
- supports: Prime Intellect
Built from the ground up for agentic post-training, Lab unifies the Environments Hub with Hosted Training and Hosted Evaluations into a full-stack platform for research and model optimization.
On 13 October 2026 Prime Intellect announced that it had acquired Moondream, whose founders Vik Korrapati and Jay Allen join to lead and scale inference and multimodal work.
Supported
Project announcement
Evidence (1): 1 supporting
- supports: Prime Intellect
We’ve acquired Moondream. Founders Vik Korrapati and Jay Allen are joining Prime Intellect to lead and scale inference, and to accelerate our work on multimodal models, including vision and computer use.
In April 2024 Fortune and The Block described Prime Intellect as a decentralized AI platform that aggregated global GPU supply for distributed model training and planned a decentralized protocol letting users own parts of collaboratively trained models.
Supported
Evidence (2): 2 supporting
- supports: The Block
The firm then wants to launch a decentralized protocol letting users own parts of the AI models they helped collaboratively train.
- supports: Fortune
The platform, currently live in beta, aggregates surplus global compute power to build AI models using distributed training across clusters.
Prime Intellect's website describes it as "The Open Superintelligence Stack" and says users can train and deploy custom AI models on an integrated compute, training and inference stack.
Supported
Project announcement
Evidence (2): 2 supporting
- supports: Prime Intellect
Train and deploy custom AI models on an integrated compute, training, and inference stack.
- supports: Prime Intellect
Prime Intellect - The Open Superintelligence Stack
Back to sectionsFunding
On 28 February 2025 Prime Intellect announced it had raised $15M led by Founders Fund, with Menlo Ventures and angels including Andrej Karpathy, bringing total funding to over $20M.
Supported
Project announcement
Evidence (3): 3 supporting
- supports: Prime Intellect
Today, we're announcing that we've raised $15M in funding, led by Founders Fund, with participation from Menlo Ventures
- supports: Prime Intellect
This brings our total funding to over $20M
- supports: Prime Intellect
Andrej Karpathy (EurekaAI, Tesla, OpenAI)
Prime Intellect's Series A post names angel investors including John Schulman, Karim Atiyeh, Aaron Levie and Dwarkesh Patel.
Supported
Project announcement
Evidence (1): 1 supporting
- supports: Prime Intellect
John Schulman (Thinking Machines), Karim Atiyeh (Ramp), Aaron Levie (Box), Dwarkesh Patel
In April 2024 Prime Intellect announced a $5.5 million seed round led by Distributed Global and CoinFund, with Compound, Collab+Currency and Juan Benet with Protocol Labs participating (Fortune, The Block).
Supported
Evidence (2): 2 supporting
- supports: The Block
CoinFund and Distributed Global co-led the round
- supports: Fortune
Distributed Global and CoinFund led the raise, closed a few weeks ago, with participation from Compound, Collab+Currency, and Juan Benet with Protocol Labs.
On 8 July 2026 Prime Intellect announced a $130M Series A. Its own post says it was led by Radical Ventures with NVIDIA Ventures, Intel Capital and Dell Technologies Capital, bringing total funding to over $150M, and PYMNTS reported the raise.
Supported
Evidence (2): 2 supporting
- supports: Prime Intellect
Today, we're announcing that we've raised $130M, led by Radical Ventures, with participation from NVIDIA Ventures, Intel Capital, Dell Technologies Capital, and our existing investors. Bringing our total funding to over $150M to build the open superintelligence stack.
- supports: PYMNTS
Prime Intellect raised $130 million in a Series A funding round to scale its tech stack and target new problems with its “open superintelligence stack.”
Cooley, the law firm that advised Prime Intellect, states that the Series A values the company at $1 billion. Prime Intellect's own post does not state a valuation.
Unverified
Evidence (1): 1 supporting
Back to sectionsTechnical progress
Prime Intellect announced INTELLECT-1 on 29 November 2024 as the first 10B-parameter language model collaboratively trained across the globe, and said it trained it across three continents with its PRIME framework.
Supported
Project announcement
Evidence (2): 2 supporting
- supports: Prime Intellect
We're excited to release INTELLECT-1, the first 10B parameter language model collaboratively trained across the globe.
- supports: Prime Intellect
We successfully trained a 10B-parameter model across three continents with our PRIME framework
On 15 April 2025 Prime Intellect announced INTELLECT-2, a 32B-parameter globally decentralized reinforcement learning training run in which anyone could contribute compute permissionlessly.
Supported
Project announcement
Evidence (1): 1 supporting
- supports: Prime Intellect
Today we are launching INTELLECT-2: the first 32B parameter globally decentralized Reinforcement Learning training run where anyone can permissionlessly contribute their heterogeneous compute resources.
On 26 November 2025 Prime Intellect released INTELLECT-3, a 100B+ parameter Mixture-of-Experts model trained on its own reinforcement learning stack, and open-sourced the weights, frameworks, datasets and environments.
Supported
Project announcement
Evidence (2): 2 supporting
- supports: Prime Intellect
Today, we release INTELLECT-3, a 100B+ parameter Mixture-of-Experts model trained on our RL stack
- supports: Prime Intellect
Our complete recipe — from the model weights and training frameworks, to our datasets, RL environments, and evaluations — has been open-sourced
Prime Intellect's GitHub organization lists public repositories including prime-rl (agentic RL training at scale), prime (a CLI and Python SDK for compute, sandboxes and environments) and prime-agent.
Supported
Project announcement
Evidence (2): 2 supporting
Whether the peer-to-peer protocol and testnet announced in February 2025 remain active is not yet verifiable. The 2026 homepage, documentation and Series A post do not mention them.
Not yet verifiable
Evidence (2): 2 context
On 14 February 2025 Prime Intellect announced a protocol and testnet, described as a peer-to-peer compute and intelligence network, and CEO Vincent Weisser reposted it.
Supported
Project announcement
Evidence (2): 2 supporting
Back to sectionsNothing on this site is financial advice.