Loading post…
Together AIAnnouncement
Announcing our $800M Series C to accelerate the shift to open-source AI
Summary
Together AI announces an $800M Series C to accelerate its open-source AI platform as production demand for inference grows. The company says the round includes $800 million from investors including Aramco Ventures, NVIDIA, Vista Equity, General Catalyst, and others, plus commitments for more than 500 MW of compute capacity to be capitalized independently by new investors. Together argues that closed frontier-model economics become unsustainable as agents and workflows scale, while open-weights models such as DeepSeek, Nemotron, MiniMax, Kimi, and GLM can provide equal or better performance at lower cost. Its full-stack approach spans models, kernels, compilers, inference systems, training infrastructure, hardware utilization, and software, with cited work including FlashAttention-4, Together Megakernel, together.compile, and expanded post-training APIs for tool calling, reasoning, and vision-language models.
Context
Together AI frames production AI economics as a full-stack problem. As agents and automated workflows generate more inference demand, the company says closed frontier models can become too expensive, while open-weights models offer customization and reported cost reductions of 6x to 20x with equal or better performance.
Approach / What changed
The company combines open and custom models with optimization across the inference stack, including kernels, compilers, inference systems, training infrastructure, and hardware utilization. Its cited work includes FlashAttention-4 for NVIDIA Blackwell, Together Megakernel, together.compile, faster endpoints for open-source models, and expanded post-training APIs for tool calling, reasoning, and vision-language models.
Takeaways
Together AI says companies using open models routinely achieve 6x to 20x lower costs while maintaining equal or better performance; Decagon reduced inference costs sixfold after moving to the platform.
The announced financing includes $800 million in equity funding and commitments for more than 500 MW of compute capacity, with that capacity to be capitalized independently by new investors.
The platform’s production strategy spans models, kernels, compilers, inference systems, training infrastructure, hardware utilization, and software rather than relying on model selection alone.
FAQ
Why does Together AI say open models can be more economical in production?
The company says closed frontier model costs that seem manageable for prototypes can become unsustainable as inference usage grows. It reports that open-weights models can provide equal or better performance at 6x to 20x lower cost, while also allowing customization and fine-tuning.
What will the new compute commitments support?
Together AI says commitments for more than 500 MW of compute capacity will support expected compute growth in the coming years. The capacity will be capitalized independently by the new investors.
What technologies does Together AI cite as part of its full-stack strategy?
The company cites FlashAttention-4 for NVIDIA Blackwell, Together Megakernel, together.compile, faster endpoints for leading open-source models, and expanded post-training APIs. The APIs support tool calling, reasoning, and vision-language models.
Source and freshness
- Source
- together.ai
- Published
- Added to Yomu
Summaries are generated from the original by a language model. Check the original before relying on it.
Keep reading
Related reading
The Good Stuff
Paid picks we use ourselves. Links carry a referral code — a subscription through one earns us a commission at no extra cost to you.
The best engineering reads, once a week
One useful email. Unsubscribe anytime.