Gradient Ventures

When Gradient Ventures launched in 2017, even other venture investors were skeptical. An AI-focused fund inside Google? Most VCs dismissed it as a corporate experiment. Eight years later, Gradient has closed its fifth fund at $220 million, managing nearly $1.2 billion in total assets, and has seeded category-defining AI companies that have produced multiple billion-dollar acquisitions.

The firm operates as a fully independent, standalone venture capital firm—spun out from Alphabet in 2025—while retaining the deep technical relationships and resources that made it distinctive. Unlike corporate venture arms that invest from a balance sheet, Gradient raised its latest fund from a diversified LP base including institutions and family offices, giving it the autonomy that comes with true independence.

Gradient's portfolio spans more than 185 companies, including some of the most consequential AI companies of the past decade. The firm led the seed round in CentML, a GPU optimization company that NVIDIA acquired in 2024. It backed Streamlit before Snowflake bought it in 2022. Lambda started as a GPU cloud provider backed by Gradient and has grown into one of the largest AI infrastructure companies in the market.

For founders building in AI, Gradient offers something most investors cannot: a direct line to the engineers and researchers building the foundational models and infrastructure that will shape the next decade of computing. The firm's partners include former founders and technical experts who have shipped AI products at scale.

Gradient's investment thesis is explicit and AI-exclusive. The firm does not invest outside of artificial intelligence and machine learning. Within that mandate, Gradient focuses on five core areas: AI-native B2B software built from day one on AI foundations, AI agents reshaping how work gets done, the AI developer stack powering the next generation of builders, real-world AI applications in biology, materials, and robotics, and open source, small models, and hybrid compute architectures.

The firm seeks companies where AI is not an add-on feature but the core architectural decision. Gradient's partners describe their ideal investment as a company that could not exist without the recent advances in foundation models—problems that became tractable only in the last few years.

Unlike generalist seed funds that spread across sectors, Gradient's portfolio is concentrated in applied AI. The firm sees AI as the most significant platform shift since mobile and is positioning itself as the earliest institutional partner for founders building at the frontier of that shift.

The firm typically leads or co-leads seed rounds and provides meaningful support in board governance, hiring, and follow-on fundraising. Gradient does not invest in ideas—they invest in companies where the founding team has already demonstrated technical execution and early traction.

In March 2026, Gradient closed Fund V at $220 million, its largest fund to date. The firm noted it saw roughly 100 companies per year that fit its thesis from 2017 to 2021—a number that has grown substantially as AI adoption has accelerated across every sector.

Recent portfolio activity shows Gradient continuing to back AI infrastructure and applied AI companies. The firm participated in the Series A for Enzo Health in May 2026 and has made at least 14 new investments in the past 12 months according to Tracxn data.

The AI infrastructure sector has been particularly active for Gradient. Companies like Lambda, which started as a Gradient seed investment, have raised later rounds with Gradient participating. The firm has also been active in AI agent companies and developer tooling—segments it views as foundational to the next wave of enterprise AI adoption.

Gradient has also supported its portfolio through multiple financing rounds, demonstrating commitment beyond the initial check. The firm's 0.37 follow-on index indicates meaningful participation in later rounds for successful portfolio companies.

Gradient's portfolio includes Lambda, a GPU cloud provider that has become one of the largest AI infrastructure companies in the market, having raised $480 million across multiple rounds. Lambda started as a hardware company and pivoted into GPU cloud services as AI demand exploded—Gradient was there at the beginning.

Writer is an AI writing and analytics platform founded by May and Waseem. The company has grown into a serious enterprise offering, competing with legacy software vendors on the strength of its AI-native architecture. Gradient led Writer's seed round and has supported the company through subsequent growth.

Streamlit, founded by Amanda, Adrien, and Thiago, was acquired by Snowflake in 2022. The company's open-source framework for building data and ML applications became a standard tool in the AI developer ecosystem. Gradient's seed investment in Streamlit produced one of the firm's most recognized exits.

CentML, founded by Anand, Akbar, Gennady, and Shang, was acquired by NVIDIA in 2024. The company built GPU optimization technology that became increasingly critical as AI training workloads scaled. Gradient led CentML's seed round, an investment that produced a strategic acquisition by the dominant GPU manufacturer.

Prepared, founded by Michael, Dylan, and Neal, is an AI-powered platform for first responders. Range, founded by Fahad and David, is a management platform for venture capital and private equity firms. Both represent Gradient's thesis of AI transforming workflows in specific vertical industries.

Other notable Gradient portfolio companies include Oura (AI health wearables), Labelbox (AI training data infrastructure), Gigs (developer hiring platform), Rad AI (radiology AI), Syrup (revenue intelligence), and Unit21 (financial crime compliance).

Gradient evaluates founders on technical rigor above all else. The firm's partners—many of whom are former engineers and researchers—want to see that founders have deep understanding of the technical problems they are solving, not just business acumen.

Obsession with quality is a consistent theme in Gradient's public communications. The firm has backed founders who insist on shipping products they consider exceptional rather than moving fast and iterating. This is particularly true in AI infrastructure and developer tooling, where technical debt is expensive to pay down later.

Speed to iterate faster than the market can copy is a specific phrase Gradient uses to describe its ideal founder. The firm believes that in AI, the window between product-market fit and competition is shorter than in previous software cycles—so execution velocity matters enormously.

Founders who want to build iconic and durable businesses, not just acquirable point solutions, resonate with Gradient. The firm explicitly looks for entrepreneurs thinking about the long arc of their company rather than an early exit.

Gradient does not require founders to have prior startup experience, but they do expect technical founders to have demonstrated the ability to ship. Academic pedigree matters less than evidence that a founder can turn ideas into working software.

Gradient's website at gradient.com is the primary channel for cold submissions. The firm reviews every submission but is selective. The highest-signal pathway remains warm introductions from Gradient portfolio founders, other institutional investors Gradient respects, or attorneys who work in the AI startup ecosystem.

Before reaching out, founders should ensure their company is clearly within Gradient's AI mandate. The firm does not invest in consumer apps, marketplaces, or non-AI enterprise software. A cold email that demonstrates clear alignment with Gradient's stated thesis will receive more attention than one that assumes broad interest.

Gradient's due diligence process typically takes two to four weeks from initial meeting to term sheet, though timing varies based on deal complexity and partner bandwidth. The firm prefers to move quickly on opportunities that fit the thesis clearly.

Founders should prepare to demonstrate technical depth in the first meeting. Gradient's partners ask hard questions about architecture decisions, model choices, and technical differentiation. Business-level pitch decks without technical substance will not advance in the process.

Following up after meetings is expected but should be spaced appropriately. Gradient partners are managing dozens of active conversations at any given time. Monthly updates on milestones achieved are more effective than weekly check-ins.

Fund V, closed at $220 million in March 2026, is Gradient's largest fund to date. The firm now manages nearly $1.2 billion in assets across five discrete funds, a significant increase from its early funds when the venture market did not fully appreciate the AI opportunity Gradient was pursuing.

At the seed stage, Gradient's average investment is approximately $5.58 million according to Tracxn data, with significant flexibility up to around $10 million for the right opportunity. The firm prefers to lead or co-lead rounds, which gives them board visibility and the ability to support follow-on fundraising actively.

The firm's portfolio companies have benefited from Gradient's willingness to participate in later rounds. With a 0.37 follow-on index, Gradient is not the most active follow-on investor, but it supports companies that demonstrate continued progress against their thesis.

Gradient's spin-out from Alphabet in 2025 means it no longer writes checks from a corporate balance sheet. The firm raised Fund V from external LPs including institutions and family offices, which gives it the independence to make investment decisions without corporate approval processes.

Founders pitching Gradient should have investor-ready financials ready to present. While Gradient is technically focused, the firm still expects founders to have a clear model for how capital turns into revenue. This means understanding your burn rate, runway, customer acquisition cost, and lifetime value.

Working with a fractional CFO is one of the most effective ways to prepare for Gradient's diligence process. A fractional CFO can build the financial models Gradient will expect, help you construct realistic projections, and ensure you can answer detailed questions about unit economics without stumbling.

Gradient's partners will challenge your assumptions. Founders who have done rigorous work on their market sizing, competitive positioning, and path to profitability will have more credibility in the room. Surface-level projections without grounding in actual data will not survive scrutiny.

Key performance indicators should be tracked and understood before the meeting. Gradient wants to see that founders know the metrics that matter for their business and can explain trends in their performance clearly.

Exploring other AI-focused venture capital firms? Our comprehensive collection of VC firm reviews covers investors across stages and sectors, including other funds with AI-exclusive mandates.

Each review provides detailed information about investment criteria, portfolio companies, and tips for securing funding. Whether you're looking for seed-stage AI investors or growth equity firms focused on AI infrastructure, you'll find relevant insights.

Finding the right investor for your AI startup is crucial to your success. The venture capital relationship is a long-term partnership—you want an investor who understands your technical domain and can provide meaningful support beyond capital.