Together AI Raises $800M Series C for Open-Source AI Push

Together AI has announced an $800 million Series C funding round aimed at accelerating open-source AI adoption, marking a significant move in the intensifying competition between open and closed AI models. The funding, disclosed on July 1, 2026, includes backing from major investors like NVIDIA, Salesforce Ventures, and Aramco Ventures. Additionally, the company has secured over 500 MW of compute capacity commitments from its new investors to meet future demand.

The round positions Together AI as a key player in the shift towards open-source infrastructure, a sector gaining traction as companies grapple with the high costs of proprietary AI systems. According to Together AI, open models on its platform routinely deliver 6x to 20x lower costs compared to closed systems, offering an economic lifeline to enterprises scaling AI-driven operations, such as customer support automation and code generation.

Open-Source Gains Ground Amid Rising Costs

The economics of AI are at a turning point. As AI tools transition from experimental to production-grade systems, the costs of proprietary models have skyrocketed, creating budget strains for companies scaling operations. A recent report from Stanford’s 2026 AI Index highlights this trend, showing uneven economic gains tied to AI-driven productivity. Closed systems like OpenAI’s GPT models monetize via API access, with costs compounding as usage scales—forcing some firms to ration intelligence during critical growth phases.

Together AI’s platform, built on open-weight models such as DeepSeek and Nemotron, allows firms to fine-tune AI solutions at a fraction of the cost. For example, Decagon, a Together AI client, slashed its inference costs by 83% after migrating to the platform. This aligns with a broader trend documented in the “Open Source Economic Index of AI Adoption,” which found open-source models leading adoption in industries like finance and software development.

Compute Access: The Next Battleground

Beyond models, access to compute power is becoming a critical differentiator. Together AI’s $800 million raise complements commitments for massive compute capacity, mirroring moves by competitors like Reflection, which recently signed a $150 million per month compute deal with SpaceXAI. Compute resources are now viewed as a competitive moat, especially as the global AI market shifts towards hybrid and open systems.

This funding also comes as geopolitical tensions shape the AI landscape. On June 25, 2026, the Trump administration reportedly asked OpenAI to limit GPT-5.6 access to government-approved partners, emphasizing national security concerns. Meanwhile, China’s release of the open-source GLM-5.2 underscores how open models are driving global innovation but also raising security risks.

What’s Next for Together AI?

Together AI’s recent innovations include FlashAttention-4 and Together Megakernel, both designed to optimize AI inference at scale. The company claims to host the fastest endpoints for leading open-source models and is among the largest producers of AI tokens globally. With the new funding, Together AI plans to expand its research and development pipeline, hire across engineering and product teams, and scale its compute infrastructure.

Looking ahead, the company’s push for open-source AI could reshape market dynamics by offering cost-effective alternatives to closed systems. However, success will hinge on maintaining competitive performance while navigating the capital-intensive demands of compute scaling and global regulatory pressures.

As the AI arms race heats up, Together AI’s $800 million bet on open-source innovation reflects both an economic imperative and a strategic vision for a more accessible AI future.

Image source: Shutterstock