Baseten, the AI inference platform behind a new wave of category-defining applications, has closed a $300 million Series E funding round at a $5 billion valuation. IVP, CapitalG, and NVIDIA anchored the investment, which marks the company’s third fundraise in just twelve months.

The round reflects surging demand for high-performance infrastructure capable of running modern AI models reliably in production environments. Additional investors include 01A, Altimeter, Battery Ventures, BOND, BoxGroup, Blackbird Ventures, Conviction, and Greylock. With this latest capital injection, Baseten has now raised $585 million since its founding in 2019.

Powering the Applications Reshaping How People Work

Six years of focused development have positioned Baseten as the inference platform behind many AI products transforming workflows across industries. The company’s customer roster includes Cursor, OpenEvidence, Abridge, Notion, and Clay, among others building specialized models for their specific domains.

“Baseten lets us run the models we need, the way we need to run them. The performance is best-in-class, but what sets them apart is everything else: the reliability, the developer experience, the fact that they’re constantly finding ways to lower our costs. They’re a partner, not a vendor.”

Shiv Rao, Co-founder and CEO of Abridge

The platform experienced explosive growth over the past year, with inference volume increasing 100-fold as more companies moved from experimentation to production-scale AI deployments.

Inference Emerges as the Dominant AI Workload

The AI industry is shifting rapidly from training ever-larger models to deploying them at scale. Industry analysts estimate that inference will account for two-thirds of all AI compute by the end of 2026, up dramatically from one-third in 2023. This transition is accelerating as reasoning and thinking models require orders of magnitude more computational resources than previous generations.

“If cloud was the foundation that enabled the last generation of great technology companies, inference is the foundation for the next. Every breakout AI application depends on fast, reliable, and cost-effective inference. We’ve spent six years building the infrastructure to make that possible—and we’re ready for this next chapter of hundreds and then thousands of new models.”

Tuhin Srivastava, co-founder and CEO of Baseten

The investment underscores a broader market shift. Rather than relying exclusively on a handful of giant models, organizations across sectors are increasingly building custom and domain-specific models tailored to their particular use cases and customer needs.

Capital Deployment Focused on Speed, Reliability, and Developer Experience

The new funding will accelerate Baseten’s platform development across three core areas. The company plans to enhance infrastructure performance, strengthen uptime guarantees, and expand its developer tools and experience. These improvements will support customers building precise models for everything from software development to medical documentation and legal analysis.

“Our customers are building precise models for everything from elite software development to medical documentation to high-stakes legal reasoning, and they need a platform that will allow them to deliver deep specificity, expertise, and performance. This is the type of user experience — and the promise of AI — we’re dedicated to making possible.”

Tuhin Srivastava

Looking ahead, Baseten will build what it describes as an AI infrastructure platform encompassing every scenario that drives increased inference workloads. As AI embeds itself in products and workflows across industries, the company sees inference becoming one of the largest computational markets ever created.

Building Infrastructure for a Multi-Model Ecosystem

Founded in 2019 and based in San Francisco, Baseten has built its platform around a fundamental conviction: the future belongs not to a few giant models, but to thousands of specialized ones. The company’s infrastructure is designed to support this multi-model ecosystem with strong guardrails, security, and observability features that give organizations control over their deployments.

The platform’s architecture uses open runtimes rather than proprietary weights, avoids lock-in around customer models, and incorporates multi-cloud flexibility to optimize for both reliability and cost. This approach allows companies to maintain ownership of their differentiation and intellectual property while accessing enterprise-grade infrastructure.

Co-founders Tuhin Srivastava and Amir Haghighat established the company with the thesis that fast, reliable, and secure inference would become the essential link between AI capabilities and real-world business impact. That vision has been validated through a series of market inflection points, from early models like Whisper and Stable Diffusion through the Llama family and diffusion transformers, to the recent open-source surge led by DeepSeek, Qwen, and GPT-OSS.

Investor Perspective on Platform Economics and Market Position

“Baseten is quickly becoming default infrastructure,” says Sarah Guo, General Partner at Conviction. “In a world where every ambitious AI team wants to run many models and fully own its IP, Baseten gives them the freedom, reliability, and economics to do that at scale. That combination—open runtimes, multi-cloud resilience, and a deeply considered developer experience—is the new standard the best companies expect.”

The rapid fundraising pace reflects both investor confidence and market urgency. With three rounds completed in twelve months, Baseten has secured the capital needed to build what the company describes as “the most ambitious version” of its platform. As reinforcement learning enables teams to tailor models directly to specific workflows, demand for robust inference infrastructure continues to accelerate across both consumer and enterprise segments.

Origianl Creator: Ekaterina Pisareva
Original Link: https://justainews.com/companies/funding-news/ai-inference-leader-baseten-secures-300m-at-5b-valuation/
Originally Posted: Fri, 23 Jan 2026 18:03:52 +0000