Nous Research Joins NVIDIA's Nemotron Coalition, Offers Nemotron 3 Ultra Free
Nous Research joins NVIDIA's Nemotron Coalition and offers two free weeks of the 550B Nemotron 3 Ultra model via Hermes Agent

- Nous Research joins NVIDIA's Nemotron Coalition alongside H Company, NAVER Cloud, and Prime Intellect to co-develop open frontier models on DGX Cloud.
- Free two-week access to Nemotron 3 Ultra via Nous Portal and Hermes Agent, running June 4–18, 2026, powered by Nebius.
- Nemotron 3 Ultra is a 550B/55B-active MoE model scoring 48 on the Artificial Analysis Intelligence Index — the highest of any US open-weight model — at 300+ tokens/sec.
- 1M token context window and up to 5.9x faster inference than comparable open-weight models; weights, training recipes, and reward models are fully open-sourced.
- Hermes Agent is a self-improving, open-source agent framework with 180K+ GitHub stars; its learning loop makes agents 40% faster after 20+ self-created skills.
- After the free period, Nemotron 3 Ultra costs $0.68/1M input and $2.67/1M output tokens; the coalition's next milestone is the Nemotron 4 model family.
Nous Research has joined NVIDIA's Nemotron Coalition and is celebrating by giving everyone free access to NVIDIA's most powerful open model yet. The deal is simple: sign up for the Nous Portal, connect Hermes Agent, and you get two weeks of Nemotron 3 Ultra at zero cost. For a model of this caliber, that's a meaningful on-ramp.
The model behind the headline
Nemotron 3 Ultra is a 550 billion parameter (55B active) open model from NVIDIA built for long-running, agentic workflows with fast and affordable performance across hundreds of tool calls. The architecture is a hybrid Mamba-Transformer Mixture-of-Experts (MoE) design, meaning only 55B of those 550B parameters fire on any given token, keeping inference costs manageable even at this scale.
The context window is one of the standout specs. The headline specs include a 1-million-token context window and native support for multi-token prediction. Mamba-2 layers handle long-range sequence dependencies with linear time complexity, replacing a subset of the attention layers that would otherwise make 1M-token context inference prohibitively expensive. In practice, that means you can feed entire codebases, long tool histories, or research trails into a single context without losing the thread.
The numbers that matter
Nemotron 3 Ultra went live on Hugging Face on June 4, 2026. It scores 48 on the Artificial Analysis Intelligence Index , the highest of any US-built open-weight model ever released , and runs at over 300 tokens per second, three to six times faster than comparable Chinese models available through commercial APIs today.
- 550B total / 55B active parameters via MoE routing
- 1M token context (NVFP4 quantized on Blackwell), 262k in BF16
- Pricing after the free period: $0.68 per 1M input tokens and $2.67 per 1M output tokens
- Throughput: 5.9x higher inference throughput compared to GLM-5.1-754B-A40B and 4.8x compared to Kimi-K2.6-1T-A32B
- Open data release: 10M new SFT samples, 1M new RL tasks, and 15 new RL environments, with SWEBench Verified scores between 65% and 70.4% across major agent frameworks
- License: OpenMDW-1.1 (Linux Foundation permissive license), with weights, post-trained checkpoints, reward models, and training recipes all released publicly
The coalition growing around it
The NVIDIA Nemotron Coalition is a first-of-its-kind global collaboration of model builders and AI labs working to advance open, frontier-level foundation models through shared expertise, data and compute. Through the coalition, Black Forest Labs, Cursor, LangChain, Mistral AI, Perplexity, Reflection AI, Sarvam and Thinking Machines Lab are among the members bringing together their expertise to collaboratively build open frontier models.
H Company, NAVER Cloud, Nous Research, and Prime Intellect are joining the Nemotron Coalition as new members. These members will contribute unique strengths spanning data, training environments, evaluation frameworks and domain expertise to support the collaborative development of an open frontier model trained on NVIDIA DGX Cloud, which will serve as the foundation for the upcoming Nemotron 4 family.
Coalition members contribute data, evaluations, and domain expertise towards an open model trained on Nvidia's DGX Cloud. The first model built by the coalition will underpin the upcoming NVIDIA Nemotron 4 family of open models. Nous Research's induction is not just ceremonial: their Hermes Agent platform becomes a first-class deployment surface for the Nemotron family, and their data and evaluation contributions feed directly into the next generation of models.
What Hermes Agent actually is
Hermes is the self-improving AI agent built by Nous Research. It's the only agent with a built-in learning loop , it creates skills from experience, improves them during use, nudges itself to persist knowledge, searches its own past conversations, and builds a deepening model of who you are across sessions.
Hermes Agent shipped its official desktop app on June 2, 2026, bundled as a public preview at v0.15.2 with native builds for macOS, Windows, and Linux. The open-source agent has now accumulated over 180,000 GitHub stars in under four months since its February 25 launch, making it the fastest-growing open-source agent framework of 2026 by Dealroom's count.
The learning loop is the core differentiator. After every task execution, Hermes adds an evaluation layer. It assesses whether the outcome succeeded, extracts reusable reasoning patterns, and stores them as skill files (plain Markdown). Next time it encounters a similar task, it pulls the relevant skill instead of reasoning from scratch. The performance claim is specific: agents with 20+ self-created skills complete similar future tasks 40% faster than fresh instances.
How to get the free access
Nous Research has been inducted into the Nemotron Coalition of leading AI labs working with NVIDIA to advance open frontier foundation models. In honor of this, they've partnered with Nebius to provide Nemotron 3 Ultra free on Nous Portal for two weeks (June 4th – June 18th).
Getting started takes about a minute. There are two paths:
- Desktop app (recommended):
Download the Hermes Desktop installer for macOS or Windows, then open it. On first launch it finishes setting itself up, usually under a minute.
Then connect your Nous Portal account and select the
nvidia/nemotron-3-ultra:freemodel variant. - Command line (macOS/Linux/WSL2): Run the one-liner below, then run
hermes setupand follow the prompts.
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bashOnce set up, you can switch models mid-session with /model nvidia/nemotron-3-ultra:free from inside any Hermes conversation. The :free suffix is what keeps it on the no-cost tier.
The bigger picture
The timing is deliberate: DeepSeek and Kimi have dominated open-weight leaderboards for most of 2026, and Nemotron 3 Ultra is NVIDIA's most direct challenge yet. By pooling compute resources, data, and research expertise, the Nemotron Coalition aims to challenge the industry's reliance on closed, proprietary ecosystems. For the AI community, this represents a major move toward transparency, sovereignty, and the democratization of high-performance intelligence.
Agent platforms including BlackBox AI, Cline, Factory AI, Hermes Agent, Kilo Code, LangChain Deep Agents, OpenClaw, OpenCode, OpenHands and Pi support the new Nemotron models. The breadth of that list signals that NVIDIA is treating Nemotron 3 Ultra as infrastructure, not just a model release. Nous Research's Hermes Agent is now one of the primary surfaces where developers can put that infrastructure to work, and the free two-week window is the lowest-friction way to find out if it fits your stack.