companies

TabbyML

Maker of Tabby, a self-hosted AI coding assistant teams run on their own infrastructure as an open alternative to GitHub Copilot.

TabbyML was founded in 2023 by Meng Zhang and Lucy Gao, both former Google engineers who had worked on computer vision and generative model teams before starting the company. Their pitch to engineering teams was straightforward: many organizations, especially in finance, defense and other regulated industries, cannot send proprietary code to a third-party API, so TabbyML built Tabby as a coding assistant that runs entirely inside a customer's own network on consumer or datacenter GPUs, with no telemetry sent back to TabbyML by default. The company raised roughly $7.2 million in early funding, including a $3.2 million round covered by TechCrunch in late 2023, and has grown the open-source project to more than 30,000 GitHub stars. Rather than training its own foundation models, TabbyML partners with providers including Mistral AI, DeepSeek and Qwen to give customers a choice of open-weight models to run behind their own firewall. TabbyML is a small, engineering-heavy team based in the San Francisco Bay Area, and continues to position itself against much larger, better-funded assistants primarily on the strength of its self-hosting and data-control story rather than trying to out-feature them on capability alone.

Founded

2023

Headquarters

San Francisco, United States

Sector

applications