Together AI raises $800M at $8.3B
Together AI closes an $800M Series C at an $8.3B valuation, with NVIDIA and Aramco Ventures among backers and 500+ MW of compute commitments attached.
Together AI announced an $800M Series C on July 1 at an $8.3 billion post-money valuation, up from $3.3 billion in early 2025. Backers include NVIDIA, Aramco Ventures, Vista Equity, General Catalyst, Salesforce Ventures, and Lux Capital. Alongside the equity, the company says new investors will independently capitalize commitments for more than 500 MW of compute capacity.
Together runs an AI-optimized cloud for open and specialized models: it serves DeepSeek, Nemotron, MiniMax, Kimi, and GLM, and builds inference tooling like FlashAttention-4 for NVIDIA’s Blackwell GPUs. Customers named in the announcement include Cognition, Cursor, ElevenLabs, Decagon, and Suno, which is a fair snapshot of who actually buys open-model inference at scale. The company claims customers see 6x to 20x lower costs than closed-model pricing for comparable quality.
The round says something simple about where the market is heading. Serving open models cheaply is now an infrastructure business large enough to attract sovereign-linked capital and a GPU vendor onto the same cap table.
If you route work across providers, this is the layer that keeps your fallback options alive. US developers are already turning to cheaper Chinese models, and briefs like Meituan’s 1.6T open coding model only matter to your stack if someone hosts them reliably. The vendor lock-in guide covers how to keep that exit ramp real.