Baseten Embedded in Benchling Inference

This partnership matters because it gives Baseten a way into biotech through the software scientists already use all day, instead of asking labs to adopt a separate AI infrastructure vendor. Benchling sits inside experiment design, sample tracking, and research records, so Baseten can power model inference exactly where proprietary biology data already lives. That makes private deployment and compliance features a buying requirement, not an add on.1 2 3 4

• Benchling is a strong distribution wedge. Benchling says Model Hub gives scientists direct access to models tied to their experimental record, and its customer base spans 1,200 plus biotechs and 200,000 plus scientists. That turns one partnership into reach across a large installed base.2 3 5
• The workflow is concrete. Scientists can run scientific AI models from inside Benchling, while Benchling Inference uses Baseten to provide the GPU layer underneath. That means the user experience stays inside Benchling, while Baseten supplies the hard part, which is serving models reliably and at scale.1 2 6
• This fits where biotech spend is moving. Benchling grew to an estimated $210M ARR by mid 2024 by becoming system of record software for biotech R&D, and its newer AI products are extending that position into model access. Baseten benefits because inference becomes embedded in an existing lab software budget and workflow.4 5 2

The next step is for inference vendors to win regulated verticals by becoming infrastructure inside category leaders, not by selling raw compute alone. If Benchling keeps becoming the control plane for scientific AI, Baseten can expand from one partnership into a repeatable model for pharma, genomics, and other data sensitive enterprise markets.1 2 4 6