Arcee AI Reaches $1B+ Valuation After Building Frontier Open-Weight Models for Just $20M

Arcee AI Secures Series B Funding at Over $1 Billion Valuation

Arcee AI has raised Series B funding that values the San Francisco-based AI model developer at more than $1 billion, according to its announcement on September 16. Vista Equity Partners, Cambium Capital and Emergence Capital led the round, with participation from A10 Ventures, Hitachi, IAG, Microsoft's M12, P7 and Wipro. While the company did not disclose the exact amount, a source told Fortune the round was at least $150M, with the valuation reaching $1 billion before the new money.1

1

2

Source: SiliconANGLE

Building Frontier Models with Unprecedented Cost Efficiency

What sets Arcee AI apart is the remarkable cost efficiency with which it developed its Trinity family of models. The company built its entire 2025 model lineup, including Trinity Large with 400 billion parameters and 13 billion active for each token, for approximately $20M. This figure covers salaries, compute, data, infrastructure and operations—a fraction of what most closed-source competitors spend on similar frontier models.1

1

Co-founder and Chief Executive Mark McQuade told Fortune that Arcee had $30M in the bank when it decided to build its own models. "I said 'let's do it' and I bet the company on it," McQuade explained, noting the company spent 65% to 70% of its capital on the effort. Before developing its own models, Arcee focused on post-training and fine-tuning services, adapting existing models for enterprise use. McQuade founded the company in 2023 after leaving Hugging Face, which Nvidia is currently acquiring.1

1

2

Competing in the Open-Weight AI Race Against China

Arcee AI is pushing to become one of the top U.S. developers of open-weight AI models, which publicly release their core numerical parameters, allowing anyone to download, modify and run them on their own infrastructure. McQuade told Fortune his goal is to catch up with Chinese open-weight models like GLM Flash from Beijing-based Z.ai. "The U.S. is far ahead in closed-source, but it kind of dropped the ball on open-source," he said.1

1

2

This approach contrasts sharply with most U.S. AI developers who have opted for proprietary models, keeping underlying code private and often dictating where those models can run. Meanwhile, Chinese AI developers have embraced open-weights, developing powerful and accessible models that match capabilities of advanced systems built by OpenAI, Anthropic and Google. "Organizations should not have to choose between the capabilities of a frontier model and the ability to control the technology at the center of their work," McQuade emphasized.2

2

Related Stories

Together AI raises $800M at $8.3B valuation as companies shift to cost-effective AI models

Cerebras Systems Raises $1.1 Billion in Pre-IPO Funding Round, Challenging Nvidia in AI Chip Market

Together AI Secures $305 Million in Series B Funding to Expand AI Acceleration Cloud

Intel spinout Articul8 raises over half of $70M funding round at $500M valuation for enterprise AI

Strategic Investment and Future Plans

Monti Saroya, senior managing director at Vista Equity Partners, praised Arcee's execution: "Arcee has demonstrated a rare combination of technical execution and capital efficiency, building frontier open-weight models at a pace and cost profile that stands out in the market. As enterprises increasingly look for AI systems they can control, customize and deploy on their own terms, we believe Arcee is building critical infrastructure."1

1

2

The company will deploy the Series B funding across three strategic areas. First, training the next generation of Trinity models, which is already underway. Second, expanding its collaboration with the U.S. Department of Energy and its 17 national laboratories on Genesis-Science-1, an open scientific computing model for research tasks. Third, building customization tools and deployment tools that help organizations test and run open models. Arcee will also work extensively with Vista's portfolio companies.1

1

2

Source: The Next Web

A day before the funding announcement, software company CIQ revealed that Arcee had chosen its Fuzzball tool to coordinate computing jobs for its next flagship model, though Arcee maintains its own training and inference software. The timing follows August's $1.1 billion raise by River AI to democratize AI by letting companies train and keep their own models, signaling growing investor appetite for alternatives to proprietary models.1

1