cd /news/ai-infrastructure/ai-infrastructure-startup-fireworks-… · home topics ai-infrastructure article
[ARTICLE · art-63996] src=siliconangle.com ↗ pub= topic=ai-infrastructure verified=true sentiment=↑ positive

AI infrastructure startup Fireworks closes $1.5B round at $17.5B valuation

Fireworks AI Inc., an AI infrastructure startup, raised $1.5 billion in Series D funding led by Atreides Management, Index Ventures, and TCV, valuing the company at $17.5 billion. The company, which processes over 40 trillion tokens daily for clients like Samsung and GitLab, will use the funds to expand infrastructure and hire engineers.

read3 min views1 publishedJul 16, 2026
AI infrastructure startup Fireworks closes $1.5B round at $17.5B valuation
Image: Siliconangle (auto-discovered)

AI infrastructure startup Fireworks closes $1.5B round at $17.5B valuation

Fireworks AI Inc., a startup that helps developers train and run artificial intelligence models, has raised $1.5 billion in funding.

Atreides Management, Index Ventures and TCV jointly led the Series D round. Fireworks stated in its funding announcement Wednesday that they were joined by more than a half-dozen other backers, including Nvidia Corp. The company is now worth $17.5 billion.

The steep valuation is partly a reflection of Fireworks’ rapid sales growth, which recently propelled its annualized revenue past the $1 billion mark. The company’s customer base includes Samsung Electronics Co., GitLab Inc. and other major tech firms. Fireworks says that it processes more than 40 trillion tokens per day for its users.

Fireworks operates a cloud platform that developers can use to fine-tune, or customize, open-source AI models. It provides access to managed graphics card clusters under a usage-based pricing model. There’s also an AI agent that automates the AI training workflow. Developers can describe the task for which they wish to fine-tune a model, upload a training dataset and have the agent sort out the technical details.

An AI model includes settings called hyperparameters that define key features such as the number of artificial neurons it contains. According to Fireworks, its AI agent finds the specific combination of hyperparameters that maximizes a model’s output quality. When needed, it can also extend the user-provided training dataset with DPO files. Those are datasets that contain instructions on how a model should answer prompts.

Developers speed up training runs by splitting calculations across multiple chips. The way training parallelization is implemented varies from project to project. According to Fireworks, its platform supports four different parallelization techniques that are each optimized for a different model type. Software teams can run all four side-by-side or use only one method at a time.

After customers fine-tune a model, they can host it using one of the two inference services offered by Fireworks. The first provides serverless environments that don’t require users to configure the underlying infrastructure. The company’s other inference service, Deployments, offers dedicated graphics card clusters. Fireworks says that they provide better performance than its serverless offering.

The Deployments interface also offers more customization options. Developers can adjust how the service performs autoscaling, the task of adding or removing hardware capacity when request volumes change. The service reduces models’ infrastructure requirements by compressing them using a method called quantization.

“Every company holds knowledge no one else has: its data, its workflows, its customers, its definition of quality,” said co-founder and Chief Executive Officer Lin Qiao. “Fireworks turns that knowledge into specialized intelligence they own and can keep improving.”

Fireworks will use the proceeds from its funding round to expand its infrastructure and hire more engineers.

Photo: Fireworks

Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.

15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more** 11.4k+ theCUBE alumni**— Connect with more than 11,400 tech and business leaders shaping the future through a unique trusted-based network.

About SiliconANGLE Media

SiliconANGLE,

theCUBE Network,

theCUBE Research,

CUBE365,

theCUBE AIand theCUBE SuperStudios — with flagship locations in Silicon Valley and the New York Stock Exchange — SiliconANGLE Media operates at the intersection of media, technology and AI.

Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.

── more in #ai-infrastructure 4 stories · sorted by recency
── more on @fireworks ai inc. 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/ai-infrastructure-st…] indexed:0 read:3min 2026-07-16 ·