Join our daily and weekly newsletter for the latest updates and exclusive content on industry-leading AI coverage. learn more
Nous Research is a New York-based AI Collective known for developing what is known as “personalized, unlimited” language models, and has launched a new inference API that makes the model more accessible to developers and researchers through the program’s interface.
The launch of the API represents a significant expansion of Nous Research’s product. This has attracted attention as it challenges the more restricted approaches of large AI companies like Openai and humanity.
“We listened to your feedback and built a simple system to make our language model more accessible to developers and researchers everywhere,” the company announced on social media.
The first API release includes two of the company’s flagship models. It features the Hermes 3 Llama 70B, a powerful general purpose model based on Meta’s Llama 3.1 architecture, and the Deephermes-3 8B preview, a user-released inference model.
Today we are releasing an inference API that serves Nous Research Models. We listened to your feedback and built a simple system to make our language model more accessible to developers and researchers everywhere.
The first release comes in two models: the Hermes 3 llama 70b and…pic.twitter.com/daea8donln.
– Nous Research (@nousResearch) March 12, 2025
Within Nous Research’s waitlist-based portal: how AI startups manage high demand
To manage demand, Nous has implemented a waitlist system through a new portal, granting access on a first-come, first-served basis. The company offers $5 free credits on all new accounts. Developers can access the API documentation to see more details about integration options.
The WaitList approach provides important insights into Nous Research’s strategic positioning. Unlike major players with large GPU reserves, Nous faces infrastructure constraints common to small AI organizations. The waitlist acts as both technical necessity and marketing tactics, creating exclusivity that generates buzz while managing computational load.
What makes this approach particularly stand out is how it reflects the grassroots spirit of Nous. The company positions it as a major alternative to technology AI, but also employs practical business strategies that acknowledge the reality of scaling inference services. This tension between idealism and practicality could define Noos’ journey, as it shifts from purely open source releases to commercial goods.
The API follows OpenAI API design patterns for completion and chat completion, making it potentially easier for developers familiar with its interface to integrate Nous’ models into their applications.
From Github downloads to cloud APIs: Evolution of Nous Research shows a new business model
The launch of the API was just four months after Nous debuted, and Nous Chat, the company’s first user-friendly chatbot interface. The company has released numerous open source models for local deployments, but the new API allows developers to access high-performance versions of these models without having to manage their own infrastructure.
“Previously, if researchers and users actually wanted to deploy these models, they had to download and run the code on their own machines. This is a time-consuming, cumbersome, and potentially expensive effort.”
Released last month, Deephermes-3 represents the company’s entry into an increasingly competitive field of inference-focused AI models. This model allows users to switch between concise responses and detailed inference processes through a system prompt that activates the “think” feature.
“Unlimited AI” Philosophy: How Nous Research challenges Big Tech’s Guardrail
Since its founding in 2023, Nous Research has been positioned as an alternative to more tightly controlled AI systems. The company emphasizes the collaboration between individual agents and users’ needs, reflected in blog posts with titles such as “Freedom at the Frontier” and “From black boxes to glass homes: Essentials for transparent AI development.”
“Superintelligence needs to be resolved for the greatest individual agency and freedom of mind,” the company wrote in a recent blog post, which announced the Psyche Project in Solana. “The development cannot be left alone in the hands of a small number of companies and the oligarchs.”
This philosophical stance resonated with developers looking for more flexible AI systems, but this approach also raises questions about responsible deployment. Despite marketing itself as “unlimited”, the company’s models include several guardrails against harmful produce.
Moneying Open AI Research: Nous’ API Strategies and Roadmap, including Hermes, DeepHelm
The launch of the API shows Nous Research’s move towards a more sustainable business model while maintaining its commitment to open source principles. According to the company’s release timeline, Nous has released 29 AI artifacts since July 2023, including models, papers, code and datasets.
The API represents a delicate but important evolution in Nous Research’s business model. By continuing to release the weights of the model and commercializing the deployment, Nous is trying to turn the difficult circle into four-sided rings. It generates revenue without alienating the open source community that forms the basis of that.
This hybrid approach appears to be designed to capture different segments of the market. Individual developers and researchers can download and run the model locally, but companies looking for reliability, convenience and performance optimization can pay for API access. In fact, Nous monetizes the infrastructure and optimization layer, not the model itself. This is a strategy that addresses the fundamental economic challenges of open source AI without undermining core principles.
The success of this approach could potentially determine whether independent AI labs can establish sustainable business models that maintain independence from large tech or venture capital companies that could drive more aggressive commercialization. For developers concerned about AI centralization, Nous’ experiments represent a potential intermediate pathway that can maintain diversity in the AI ecosystem.
Nous’s research shows that the provision of that inference expands over time, potentially allowing more models like Hermes 2 Pro to call features or specialize in its mental projects.
For an increasing ecosystem of AI startups based on open models, the new API offers another option beyond established players such as AI, humanity, Openai, and more, intensifying competition and fostering further innovation in the AI reasoning space.
“We welcome your ideas to help shape the future,” the company said in its announcement, further highlighting its community-oriented approach to AI development.
