Prime Intellect Stock
| Valuation | $1B1 |
| Valuation Date | Jul 8, 2026 |
| Implied Valuation | $1.66B |
| Last Price | $8.29 |
| 52-Week High | $8.29 |
| 52-Week Low | $5.00 |
| YTD Return | n/a |
| 1-Year Return | n/a |
| 3-Year Return | n/a |
| Total Funding | $150M2 |
| Latest Round Amount | $130M1 |
| Latest Round | Series A |
| Funding Rounds | 3 |
| IPO Status | n/a |
| Expected IPO Date | n/a |
| Revenue | $100M2 |
| Revenue Type | Run Rate |
| P/S Ratio | 10.0 |
About Prime Intellect
Prime Intellect is an artificial intelligence company that provides infrastructure for organizations to train, deploy, and continuously improve their own AI models and agents based on open-source LLMs. Its platform bundles GPU compute, a large-scale reinforcement learning framework, hosted training, code-execution sandboxes, evaluations and inference into what the company calls an "open superintelligence stack." It also runs an open research lab that releases frontier models such as its INTELLECT series, trained across globally distributed hardware.
Invest in Prime Intellect
Prime Intellect is a private company, so you can't purchase it with a regular brokerage account. However, its shares are available on Clarity (formerly Hiive).
Clarity is a secondary marketplace where accredited investors can buy and sell shares of private, pre-IPO companies. Invest in the world's most exciting startups before they go public.
Invest on ClarityStock Analysis may be compensated for user sign-ups through Clarity links. Clarity is not affiliated with Stock Analysis. Securities offered by Clarity Capital Limited, member of FINRA/SIPC.
Related Companies
News
Moondream is joining Prime Intellect
Moondream confirmed it has been acquired by Prime Intellect. Moondream Cloud inference and Lens will shut down on November 13, 2026, with unspent credits refunded.
Prime Intellect acquires Moondream, a vision AI startup founded in Seattle by AWS vets
GeekWire reported Prime Intellect's acquisition of open-source vision AI startup Moondream, citing The Information; terms were not disclosed and Moondream had fewer than 10 employees.
Neocloud Prime Intellect Acquires Startup To Speed Up Inference
The Information first reported that Prime Intellect acquired Moondream, a small vision-model startup, to speed up open-model inference. Terms were not disclosed.
Moondream is joining Prime Intellect
Prime Intellect acquired Moondream on Oct. 13, 2026; founders Vik Korrapati and Jay Allen joined to lead inference and multimodal model work, integrating the Photon engine into Prime's stack.
Prime Inference: Fast, Reliable Serving for Frontier Open Models
Prime Intellect launched Prime Inference, offering serverless and reserved GPU serving for frontier open-source models across multiple datacenters, including a public GLM-5.3 deployment.
Prime Sandboxes: MicroVMs for Agentic RL Training at Scale
Prime Intellect launched Prime Sandboxes in general availability, offering Linux microVMs for large-scale agentic reinforcement learning with Docker support and usage-based pricing.
GLM-5.2 RL weight transfer in 4 seconds using NIXL and ModelExpress
Prime Intellect cut post-training weight transfer for GLM-5.2 from about 86 seconds to under four seconds by rebuilding trainer-to-inference sync on NIXL and ModelExpress.
Prime Intellect publishes benchmark on autonomous AI research
Prime Intellect ran a public experiment with 18 frontier models across 153 autonomous research tasks, reporting that Claude Fable 5 closed roughly 82% of the gap to the human baseline.
Prime Flash MoE - Faster MoE Kernels optimized for Blackwell
Prime Intellect released Prime Flash MoE, Blackwell-optimized CUDA kernels that fuse routing-aware GEMMs, SwiGLU, quantization, and reductions to speed mixture-of-experts inference on NVIDIA B200 hard...
Prime Agent: A self-improving RLM agent
Prime Intellect launched Prime Agent, an open-source self-improving coding harness built on Recursive Language Models and a Continual Harness that lets agents modify their own state during long-runnin...
Scaling Agentic RL: 365,000+ Environments for SWE, Terminal, and Search
Prime Intellect integrated 23 open-source agentic tasksets behind one API, offering roughly 365,000 software engineering, terminal, and search tasks for evaluation and RL training on its infrastructur...
$130M Series A to Build the Open Superintelligence Stack
The company said the round from Radical Ventures, NVIDIA Ventures, Intel Capital, and Dell Technologies Capital brings total funding above $150 million.
Prime Intellect Raises $130M Series A to Build the Open Superintelligence Stack
Intel Capital highlighted Prime Intellect's infrastructure for training and deploying open superintelligence workloads for enterprises.
Prime Intellect raises $130M to scale AI training platform
Prime Intellect secured $130 million from investors including NVentures and Intel Capital, valuing the company at $1 billion ahead of broader enterprise agent deployments.
Prime Intellect raises $130M Series A to help enterprises build their own AI agents
Prime Intellect raised a $130 million Series A at a $1 billion valuation, led by Radical Ventures with participation from Nvidia Ventures, Intel Capital and Dell Technologies Capital. The company said...
Prime Intellect Joins the NVIDIA Nemotron Coalition to Advance Open Frontier Models
Prime Intellect joined the NVIDIA Nemotron Coalition alongside Mistral AI, Perplexity, Cursor, and others, contributing its 2,500+ open RL environments and verifiers framework for post-training Nemotr...
Releasing Lab: the training platform for self-improving agents
Prime Intellect made its Lab agent-training platform generally available after private beta, supporting hosted training on models from NVIDIA, OpenAI, Meta, and Qwen.
Leveraging NVIDIA to Build the Open Superintelligence Stack
Prime Intellect detailed how NVIDIA Blackwell systems, Dynamo inference, Nemotron models, and NeMo Gym integrations power its hosted agent training and sandbox infrastructure.
INTELLECT-3: A 100B+ MoE trained with large-scale RL
Prime Intellect released INTELLECT-3, a 100-billion-parameter Mixture-of-Experts model trained on its reinforcement learning stack across a 512-GPU NVIDIA H200 cluster. The company said it reaches sta...
INTELLECT-2 Release: The First 32B Parameter Model Trained Through Globally Distributed Reinforcement Learning
Prime Intellect released INTELLECT-2, a 32-billion-parameter reasoning model trained via fully asynchronous reinforcement learning across a permissionless swarm of distributed compute contributors. Th...
%20(1)-1.png?v=intellect-1)
INTELLECT-1: Launching the First Globally-Distributed Training of a 10B Parameter Model
Prime Intellect launched the first globally distributed training run of a 10-billion-parameter language model, coordinating compute contributed across three continents. It marked an early demonstratio...
AI startup Prime Intellect raises $5.5M to build high-powered, decentralized research platform
Prime Intellect announced a $5.5 million seed round co-led by Distributed Global and CoinFund to launch a decentralized platform for collaboratively building open-source AI models. The platform aggreg...