We raised $8 Million to bring Kilo Speed to Agentic Engineering
Today we announced that Kilo Code has raised $8m in seed funding. The funding round was led by Cota Capital with participation from Breakers, General Catalyst, Quiet Capital, and Tokyo Black. This is an exciting moment for us and I want to pause and share why I am personally so energized about Kilo.
My last venture was building Brooklyn Data, a consultancy that helped companies unlock the value of their data. I started Brooklyn Data in 2018 and we rode a wave that was far bigger than we were: companies adopting cloud data warehouses like Snowflake, the modern data stack including dbt, and the rise of analytics engineering. It was an incredible moment. The foundational technology changed, the culture of data changed, and those two changes created a cascading effect that 10xed the impact of data work.
After I sold Brooklyn Data and stepped back from the company a year later, I had a nagging feeling: Would I ever again have a front row seat to such a great moment of change?
And as if right on cue, I was introduced to my cofounder Sid (co-founder, exec chair, and former CEO of GitLab) by a mutual friend. We both saw the incredible potential of AI to make every developer in the world dramatically more productive, as well as the frustrations with available AI coding tools. We had to do something about it (more on that below).
Since launching earlier this year, Kilo has gone from 0 to 750,000+ downloads, risen to #1 on OpenRouter, and is processing over 6 trillion tokens monthly. In October, we launched a Teams plan, followed two weeks later by an Enterprise plan. In the last month we have released parallel agents, one-click deploy, code review, autocomplete, managed indexing, and app builder. Thousands of developers are using these features today, and loving them.
This pace would have been unheard of even a year ago, but agentic engineering has enabled us to move faster than ever before. We call this Kilo Speed.
We’re fighting AI drag
The AI coding landscape is maturing incredibly fast. Gartner is two years into the Magic Quadrant for AI Coding Assistants. There are many companies throwing their hat in the agentic development ring—from pure plays to foundational labs and hyperscalers.
Why Kilo? We believe engineers deserve better.
With AI coding, engineers should be moving faster than ever, but tools like Cursor and GitHub Copilot are introducing AI drag, artificial friction that slows down development. Downgraded models, rate limiting, and model lock-in create a new breed of frustrations for engineers. Overage fees, limits on “premium requests”, and pricing that requires a PhD to decipher—other agentic development tools have more fine print than a used-car warranty. It ends up feeling like these tools are working against you, rather than with you.
And it’s all intentionally designed to be confusing so that you spend more money or don’t get the value of what you paid for. They’re selling gym memberships, hoping that you don’t actually show up to work out (and then charging you extra when you do).
We’re also seeing other platforms lock developers into an unnecessarily limited number of models. This seems absurd when new models launch every week; we know, we launched three last week. You shouldn’t have to choose between the latest frontier model, the ultra low latency open source model, and the extremely cost effective model from an upstart model lab. Everything in AI moves so quickly that anchoring yourself to a fixed set of models is a recipe for falling behind. Kilo has over 500 models and counting, including models from OpenAI, Anthropic, xAI, and Mistral AI. You’ll always have the best model for the job, and it never costs more than the list price of the lab.
Building for Kilo Speed
We’re bringing Kilo Speed to agentic engineering with an all-in-one platform that accelerates every single person on your team.
Never interrupt developer flow state. At Kilo, we’re obsessed with eliminating every drop of developer friction. Just like an aircraft engineer refining a prototype in a wind tunnel, our engineers have spent countless hours honing our own workflow—using Kilo to develop Kilo. And we’re incredibly proud of where we’ve landed.
With Kilo, you can kick off a session from the iOS app (reach out to get early access 😁), continue it instantly from your IDE, the CLI, or the Cloud Agent experience at your desk, have Code Reviewer give it a pass, and deploy with Kilo Deploy. Work persists across Kilo, so you can pick it up wherever you are without losing momentum. All sessions are Kilo Sessions.
Make every developer a 1000x engineer. We’re also big believers that Kilo should enhance what makes humans uniquely special, not try to replace them. This means elevating the developer from first-chair violinist to conductor of the orchestra. From code writer to visionary, architect, and manager of agents. Kilo gives every developer a fleet of parallel agents coordinated by an in-IDE Agent Manager, all grounded in a persistent Memory Bank that understands your project over time.
Accelerate AI adoption across the engineering org. We move fast at Kilo because everyone uses AI in their daily workflows—from engineering to marketing and sales. Yes, we hire team members who are passionate about AI, but we also invest in culture and tooling to support the team. Many of these features we’ve built directly into Kilo.
Shared modes, credit pooling, shared BYOK, managed indexing all started as solutions to real pain points we faced ourselves. And a personal favorite: the AI Adoption Dashboard is basically using Kilo to understand how Kilo uses Kilo to build Kilo.
Teams shouldn’t have to choose between locked-down tools, wasted tokens, or lost context across scattered systems. With Kilo, we’ve built the best of both worlds: pooled credits, centralized billing, data-privacy controls, and usage analytics so managers can adopt AI with confidence—plus collaboration tools like shared modes, so your engineers’ AI best practices become company best practices.
What’s ahead
This funding accelerates our roadmap: smarter multi-agent collaboration, enterprise-grade tooling for technical leaders, and a feature set that continues to accelerate the AI flywheel for development teams using Kilo.
The velocity of AI coding innovation is unlike anything I’ve experienced. Features that seemed impossible six months ago are shipping today, and the pace is accelerating. That’s exactly why we built Kilo the way we did. Open source. Model agnostic. Transparent pricing. No artificial limits. As the agentic coding landscape continues to rapidly evolve, Kilo users won’t be stuck waiting for their tools to catch up.
Kilo Speed isn’t just about tools, it’s a way of working. It’s about removing both AI and organizational dysfunction so teams can operate at their full potential. I’ve written more about the philosophy behind Kilo Speed here, and on Friday I’m joining Brian Turcotte from our Developer Relations team to dig deeper into how we practice what we preach. I hope you can join us.
Congratulations!
Congratulations 👏🎉