Anaconda, the Austin, Texas company behind the widely used Python distribution, said on July 15, 2026 that it had acquired Kilo Code, an open-source AI coding agent that works with many different AI models. Kilo says more than 3 million developers use it in VS Code, JetBrains editors, the web and the command line.
Terms were not disclosed. Kilo's products, pricing plans and customer support continue without immediate changes, according to Unite.AI, and Anaconda said it would announce details on how the two platforms will be combined later.
The deal is Anaconda's second acquisition of 2026. In April it bought Outerbounds, the company behind the open-source Metaflow framework first built at Netflix, which added tools for running AI workflows in production.
What Kilo Code built
Kilo is built for what the industry calls agentic engineering: AI agents that read a code repository, plan a feature, edit files, track down bugs and carry out tasks with several steps, instead of only suggesting single lines. It offers separate modes for writing code, planning a project's architecture and debugging, and has added code review, remote cloud agents and a Slack integration.
The product does not tie users to one AI lab. Its gateway gives access to more than 500 models, from commercial frontier systems to open-weight models, and can pick a model for each task automatically, for example a stronger model for architecture work and a cheaper one for routine edits. Teams can also run it on their own infrastructure.
Usage grew quickly. Anaconda said Kilo went from zero to more than 3 million developers in 16 months, mostly through word of mouth, and processes trillions of tokens a month. Unite.AI reported that Kilo now handles almost 10 trillion tokens a month. In December 2025 the company said it was processing more than 6.1 trillion tokens a month and had reached the top spot on OpenRouter.
“Kilo and Anaconda are a rare fit: almost no overlap, and what each of us lacks, the other already has.”
Why Anaconda is buying
Anaconda pitched the deal around cost and control. The company said AI token spending is rising faster than the return companies see from it, and that most AI projects never reach production. It said enterprises want agents inside the tools their developers already use, with self-hosting and a choice of models.
Anaconda says its packages, models and environments support AI development at 95% of the Fortune 500. It plans to link Kilo to that catalogue so that the same approved packages and models follow a project from the first prompt to production. The company said this deeper connection is something it is building toward and is not available yet.
Chief executive David DeSanto said every enterprise customer was asking how to let developers move at the speed AI allows without losing control of what ships or failing a compliance audit, and that Kilo is where that question begins.
The founders and the funding
Kilo was co-founded by chief executive Scott Breitenother, who earlier founded the consultancy Brooklyn Data, and Sid Sijbrandij, co-founder and executive chair of GitLab. The company is based in San Francisco and launched in early 2025.
In December 2025 Kilo raised an $8 million seed round led by Cota Capital, with Breakers, General Catalyst, Quiet Capital and Tokyo Black taking part. At the time it said the money would go to product development and partnerships with model providers.
Sijbrandij said the two companies had almost no overlap and that each had what the other lacked. He said Kilo brought a large base of developers using any model and provider, while Anaconda had spent more than a decade building trust with enterprises. Breitenother said Kilo would now build on the trust Anaconda had earned in the open-source community and in large companies.
A busy market for AI coding tools
Kilo competes in a crowded field of AI coding assistants and agents, many of them tied to a single model provider. Its pitch has been an open-source tool with no lock-in and simple access to many models.
For Anaconda, the two 2026 deals cover both ends of AI software work. Outerbounds handles running and monitoring workloads in production, and Kilo sits in the editor where code is first written. Unite.AI noted that how much value customers get will depend on how closely the products are tied together, and whether governance controls get in the way of the flexibility that drew developers to Kilo.
