According to a recent LinkedIn post from Together AI, the company is now offering NVIDIA’s Nemotron 3 Super model on its AI Native Cloud platform. The post describes Nemotron 3 Super as an open hybrid mixture-of-experts model with 120 billion parameters and 12 billion active per inference, designed for multi-agent and complex reasoning workloads.
Claim 55% Off TipRanks
- Unlock hedge fund-level data and powerful investing tools for smarter, sharper decisions
- Discover top-performing stock ideas and upgrade to a portfolio of market leaders with Smart Investor Picks
The company’s LinkedIn post highlights technical features including a hybrid Mamba-transformer architecture that is presented as delivering more than 50% faster throughput than leading open models. It also points to a 1 million token context window aimed at supporting multi-agent tasks and cross-document reasoning, with benchmark performance noted across coding, cybersecurity, finance, search, and retail use cases.
The post further indicates that Nemotron 3 Super is positioned as production-ready on Together AI’s infrastructure, referencing a 99.9% service-level objective and dedicated resources on its AI Native Cloud. From an investor perspective, expanding access to advanced NVIDIA models could enhance Together AI’s value proposition for developers, potentially driving higher usage, stickier workloads, and improved competitive positioning in the AI infrastructure and model-hosting market.
The referenced collaboration with NVIDIA AI suggests a closer alignment with a key ecosystem partner, which may support Together AI’s ability to attract enterprise customers seeking scalable agentic AI capabilities. If adoption of Nemotron 3 Super within the platform gains traction, this could translate into incremental revenue opportunities tied to higher compute consumption and differentiated, higher-margin services around complex reasoning applications.