Nous Research

Name: Nous Research
Brand: Nous Research

AI NativeLeader

World-class open source models

Est 2022|Valuation $1.5B|Raised $70M|9 Ref Customers

LinkedIn 22K|Web traffic 4.7M|FTEs 48|HQ Austin, TX

MarketsFoundation Models

01Competitive Landscape

Narrative comparison

- For teams that require open model weights and permissive experimentation, Nous Research may be relevant, particularly when Hermes behavior matches use-case requirements and open tooling is valued. - For procurement-ready enterprise inference, vendors with documented pricing, procurement workflows, and enterprise SLAs (for example, Together AI and similar providers) may be easier to evaluate. - For use cases that prioritize ultra-low latency inference for supported models, specialized providers such as Groq are commonly considered for performance-focused workloads, though model availability and quota policies should be evaluated.

Competitor perspectives (how they position relative to Nous)

Competitors that position themselves for production-grade inference emphasize documented pricing, managed infrastructure, and procurement readiness—areas where Nous’s public documentation appears less comprehensive.

	Nous Research (Portal/Forge + open weights)	Together AI	Groq
Starting Price	Not clearly published for Forge/Portal (Portal reportedly offers a small free credit bundle followed by paid credits)	Public pay-as-you-go pricing page with metered token and compute-hour options	Public pricing page (token-based offerings; model pricing varies)
Suitable For	Teams prioritizing access to open weights, experimentation with Hermes and Psyche tooling, and research use cases	Teams seeking a broad model catalog with managed endpoints and fine-tuning options	Teams prioritizing low-latency inference for supported models and predictable service costs
Key Strength	Public release of model weights and research artifacts; active research into reduced inter-GPU communication and decentralized training coordination	Commercial focus on managed endpoints, fine-tuning, and clearer procurement workflows	Focus on latency and throughput with specialized hardware and predictable pricing
Key Weakness	Public documentation gaps for enterprise security/compliance and pricing transparency; limited third-party review coverage	Perceived as a managed aggregation of model providers by some; costs depend on model and compute choices	Constraints on model selection relative to general-purpose GPU clouds; community notes rate and quota considerations

01Competitive Landscape

Narrative comparison

For teams that require open model weights and permissive experimentation, Nous Research may be relevant, particularly when Hermes behavior matches use-case requirements and open tooling is valued.
For procurement-ready enterprise inference, vendors with documented pricing, procurement workflows, and enterprise SLAs (for example, Together AI and similar providers) may be easier to evaluate.
For use cases that prioritize ultra-low latency inference for supported models, specialized providers such as Groq are commonly considered for performance-focused workloads, though model availability and quota policies should be evaluated.

Competitor perspectives (how they position relative to Nous)

	Nous Research (Portal/Forge + open weights)	Together AI	Groq
Starting Price	Not clearly published for Forge/Portal (Portal reportedly offers a small free credit bundle followed by paid credits)	Public pay-as-you-go pricing page with metered token and compute-hour options	Public pricing page (token-based offerings; model pricing varies)
Suitable For	Teams prioritizing access to open weights, experimentation with Hermes and Psyche tooling, and research use cases	Teams seeking a broad model catalog with managed endpoints and fine-tuning options	Teams prioritizing low-latency inference for supported models and predictable service costs
Key Strength	Public release of model weights and research artifacts; active research into reduced inter-GPU communication and decentralized training coordination	Commercial focus on managed endpoints, fine-tuning, and clearer procurement workflows	Focus on latency and throughput with specialized hardware and predictable pricing
Key Weakness	Public documentation gaps for enterprise security/compliance and pricing transparency; limited third-party review coverage	Perceived as a managed aggregation of model providers by some; costs depend on model and compute choices	Constraints on model selection relative to general-purpose GPU clouds; community notes rate and quota considerations