Gradient Ventures — now branded simply as Gradient — is the early-stage venture capital firm launched by Google in 2017, based in San Francisco. It was among the first venture firms anywhere dedicated to investing in AI companies at pre-seed and seed stages, and it has since grown into what it calls the seed platform of choice for AI founders. In March 2026 the firm announced Fund 5, a $220 million vehicle, taking total assets under management to more than $1 billion across five funds.

The firm's strategy is deliberately narrow and deep: it leads investments at the earliest stages of AI innovation, and its portfolio companies get structured access to Google's technical resources, engineers and AI expertise — a genuine operating edge rather than a logo on a slide deck. To date, Gradient has backed more than 500 AI founders. Notably, while Google was the firm's sole original fund investor, it is now one of many limited partners, with managing partner Zachary Bratun-Glennon noting that Gradient has added a roster of well-regarded institutional investors to the platform.

The track record at seed is striking. Gradient has led seed investments in category-defining companies including Lambda (AI cloud infrastructure), Streamlit (acquired by Snowflake), Writer (enterprise AI writing), Range, CentML (acquired by NVIDIA) and Krea (generative AI design). The firm's public values — founder-centric, collaborative, transparent and bold — translate into a hands-on operating style: Gradient describes its team as "builders, mentors, and champions" who work side by side with founders through the technical and commercial gauntlet of shipping AI products.

  • Lambda — AI compute infrastructure, a flagship seed bet
  • Streamlit — data-app framework, acquired by Snowflake
  • Writer — enterprise generative AI platform
  • CentML — ML optimisation, acquired by NVIDIA
  • Krea — generative design tools

Why it matters for founders

For an AI founder at pre-seed or seed, there is almost no better lead investor on paper: deep technical diligence, direct pipes into Google's research and infrastructure, and a portfolio signal that de-risks follow-on rounds. The caveat is selectivity — with a thesis this narrow, your moat must be genuinely AI-native, not an AI wrapper.