The open-source AI infrastructure company raised over $1 billion in its Series C, with a new blockchain partnership aimed at slashing inference costs for developers.
Together AI, a company that helps developers build and run open-source AI models, has crossed the $8 billion valuation threshold. The milestone comes after a Series C funding round that pulled in over $1 billion, more than doubling the $3.3 billion valuation the company carried just over a year ago.
The Series C round closed in April 2026 and included a $1 billion primary tranche along with a $44.89 million Series C-1 extension, landing the company at a $7.5 billion valuation on paper, though reporting indicates the effective valuation now exceeds $8 billion.
For context, Together AI’s February 2025 Series B valued the company at $3.3 billion. That means the valuation roughly 2.4x’d in about 14 months. Cumulative funding now sits at approximately $1.58 billion. The platform currently supports over 450,000 developers and touts 2-3x faster performance on open-source models compared to major cloud providers.
The crypto angle: Pearl Research Labs partnership #
On May 15, 2026, Together AI announced a strategic partnership with Pearl Research Labs, a move that introduces blockchain-based cost subsidies into Together AI’s inference pipeline.
Pearl Research Labs operates the Pearl Network, which uses a Proof-of-Useful-Work protocol. Instead of miners solving arbitrary hash puzzles, Pearl’s validators perform matrix multiplications, the same math that powers AI model training and inference. Emissions from the PRL token help offset inference costs for developers using Together AI’s platform, with the subsidy reportedly reducing costs by more than 25% for certain models.
Background: the open-source AI arms race #
Together AI has carved out its niche by specializing in open-source and open-weight AI models. While OpenAI and Anthropic keep their model weights locked behind API calls, a growing ecosystem of alternatives, led by Meta’s Llama family and Mistral’s models, has created demand for infrastructure that can run these models efficiently.
What this means for investors #
The Pearl Research Labs partnership adds a wrinkle worth watching. If Proof-of-Useful-Work protocols can meaningfully reduce AI compute costs, they create a genuine use case for crypto tokens that doesn’t rely on speculation or narrative cycles. A 25% cost reduction is the kind of concrete benefit that could drive adoption among developers who otherwise wouldn’t touch anything blockchain-related.
For the crypto-AI intersection specifically, decentralized compute networks like Akash, Render, and io.net have been pursuing similar ideas, but Together AI’s scale—450,000 developers and $1.58 billion in funding—gives the Pearl partnership a distribution advantage most crypto-native projects can only dream about. Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our