Together AI raises $305M Series B at $3.3B valuation, doubling down on open-source model inference
Feb 21, 2025
Key Points
- Together AI raises $305M Series B at $3.3B valuation, led by General Catalyst, betting that enterprises will adopt open-source models over proprietary alternatives.
- The inference layer has matured into a defensible business as open-source models gain enterprise traction, shifting infrastructure investment away from frontier model training.
- Together AI will use the capital to expand product and go-to-market as competition in model inference intensifies.
Summary
Together AI raised $305 million in Series B funding at a $3.3 billion valuation, led by General Catalyst with participation from Long Journey Ventures. The round underscores sustained investor appetite for infrastructure plays betting on open-source model adoption.
Lee Jacobs, a partner at Long Journey Ventures, announced the raise, citing belief in "open source models and their ability to empower choice for AI developers." Together AI's pitch centers on model delivery—hosting and vending open-source models into enterprises rather than locking developers into proprietary walled gardens.
The timing reflects a broader infrastructure cycle. Open-source models have matured enough that enterprises are willing to adopt them, and the inference layer has become a defensible business. Long Journey Ventures, which led the announcement, has built conviction in this thesis across multiple bets. The firm was early in Crusoe Energy and positioned early in Together AI as well—a pattern that, if executed across a few portfolio companies, can generate significant returns.
The capital gives the team substantial dry powder to expand product and go-to-market as competition in model inference intensifies. Hosts noted excitement about the product roadmap, flagging this as another major AI round focused on the delivery and distribution layer rather than frontier model training.