Sail Research Lands $80 Million

<p><strong><span class&equals;"legendSpanClass">SAN FRANCISCO<&sol;span><&sol;strong> &&num;8212&semi; Sail Research&comma; an infrastructure company for AI agents&comma; has raised &dollar;80 million in Seed and Series A funding at a &dollar;450 million valuation&period; The Series A was led by Kleiner Perkins&comma; and the Seed was led by Sequoia&period; Other participating investors included Redpoint Ventures&comma; Theory Ventures&comma; Vine Ventures&comma; CRV&comma; A&ast;&comma; and Abstract Ventures&comma; in addition to angel investors&comma; including John Hennessy&comma; chairman of Alphabet Inc&period;&comma; Lip-Bu Tan&comma; CEO of Intel&comma; and Tri Dao&comma; Chief Scientist at Together AI&period;<&sol;p>&NewLine;<p>The next chapter of AI is agents that work autonomously on complex tasks over hours and days&comma; not the brief&comma; turn-by-turn interactions that today&&num;8217&semi;s infrastructure was built for&period; While the underlying stack was optimized for responding to a human waiting at a prompt&comma; it wasn&&num;8217&semi;t built for the ways in which AI agents are fundamentally different from humans&period; Agents are limited only by the compute and context available to them&comma; and the more they&&num;8217&semi;re given&comma; the better the work they produce&period; With <a href&equals;"https&colon;&sol;&sol;edge&period;prnewswire&period;com&sol;c&sol;link&sol;&quest;t&equals;0&amp&semi;l&equals;en&amp&semi;o&equals;4719361-1&amp&semi;h&equals;2745721378&amp&semi;u&equals;https&percnt;3A&percnt;2F&percnt;2Fwww&period;gartner&period;com&percnt;2Fen&percnt;2Fnewsroom&percnt;2Fpress-releases&percnt;2F2026-1-15-gartner-says-worldwide-ai-spending-will-total-2-point-5-trillion-dollars-in-2026&amp&semi;a&equals;global&plus;AI&plus;spend&plus;projected&plus;to&plus;reach&plus;&percnt;242&period;5&plus;trillion&plus;in&plus;2026" target&equals;"&lowbar;blank" rel&equals;"nofollow noopener">global AI spend projected to reach &dollar;2&period;5 trillion in 2026<&sol;a>&comma; the most ambitious agent workloads remain out of reach for most organizations&comma; constrained not just by cost but by the rate limits and scale ceilings of platforms never designed for long-horizon use&period;<&sol;p>&NewLine;<p>Sail Research was built to remove those constraints&period; The company provides the first infrastructure platform purpose-built for long-horizon AI agents&comma; with two core components&colon; an inference stack rebuilt from the ground up around throughput and efficiency&comma; designed for agents spending billions of tokens on a single task&semi; and Sailboxes&comma; a sandbox environment built to run for hours and days that only charges for time agents are actually doing work&period; Together&comma; they give teams the economics and the scale to build agents that are maximally ambitious&comma; without hitting the walls that stop most production deployments in their tracks&period;<&sol;p>&NewLine;<p>&&num;8220&semi;Sail exists to make intelligence abundant&comma;&&num;8221&semi; said Neil Movva&comma; co-founder and CEO of Sail&period; &&num;8220&semi;Every decision we make&comma; from the chip level to the API&comma; is about giving teams the tokens&comma; the scale&comma; and the runtime to build agents without limits&period;&&num;8221&semi;<&sol;p>&NewLine;<p>Sail&&num;8217&semi;s efficiency advantage stems from a combination of proprietary infrastructure optimizations&comma; including deep customization of open-source inference engines to push GPU performance toward frontier capabilities&semi; intelligent workload distribution across providers for maximum resilience&semi; and the strategic use of underutilized compute&period; In a recent benchmark&comma; Sail&&num;8217&semi;s inference topped BrowseComp-Plus&comma; a leading deep research evaluation&comma; achieving 90&period;72&percnt; accuracy at up to 10 times lower cost than leading alternatives&period;<&sol;p>&NewLine;<p>&&num;8220&semi;Most inference infrastructure was designed to minimize latency on a single request&comma; but that&&num;8217&semi;s the wrong optimization for agents&comma; which need to sustain throughput across thousands of concurrent calls over hours&comma;&&num;8221&semi; said Samir Menon&comma; co-founder and CTO of Sail&period; &&num;8220&semi;We&&num;8217&semi;ve rebuilt the stack around that constraint&comma; and the efficiency gains compound across every layer&period;&&num;8221&semi;<&sol;p>&NewLine;<p>Co-founder and CEO Neil Movva spent years at NVIDIA pushing GPU performance to its limits&comma; then built infrastructure expertise at Apple and Together AI&period; Co-founder and CTO Samir Menon also comes from Apple&comma; where he built systems at massive scale&period;<&sol;p>&NewLine;<p>&&num;8220&semi;The infrastructure layer for the agent era is one of the most important bets in AI right now&comma; and Neil and Samir are exactly the founders to build it&comma;&&num;8221&semi; said Aditya Naganath&comma; partner at Kleiner Perkins&period; &&num;8220&semi;They bring a rare combination of deep compute expertise and systems rigor that only comes from having built at the limits of scale&period; Together&comma; they&&num;8217&semi;re building the defining inference platform for long-horizon agents&period;&&num;8221&semi;<&sol;p>&NewLine;