Open-weight models keep multiplying, and the AI coding harness around them is becoming just as important to get right. Developers and enterprises are trading off cost against capability. The harness helps translate a model’s capabilities into useful work.
Cline Bot Inc. is betting that keeping that layer open will let more people build with AI. The company’s open-source harness is growing beyond coding into a more general tool, allowing users to choose faster, cheaper inference when a task doesn’t need a frontier model, according to Renee Huang (pictured), chief executive officer of Cline.
“We don’t think AI should only be given to the people who can afford it,” Huang said. “We really think AI should belong to any type of day-to-day work.”
Huang spoke with theCUBE Research’s John Furrier and Dave Vellante at Fully Connected 2026, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed open-weight models, the AI coding harness and Cline’s growth on CoreWeave Inc.’s platform. (* Disclosure below.)
Why the AI coding harness matters for open-weight models
Cline has spent two years building a rigorous evaluation system based on real-world traffic, Huang explained. It tunes the harness to work well on both large frontier models and smaller, cheaper ones. The company recently open-sourced its evaluations for open-weight agents.
“How do we design a very modular architecture behind the scenes to give the model the right tool, the right correction loop?” Huang said. “How do they detect failure and how do they fix themselves?”
Demand for that choice is reflected in usage. Cline’s token consumption has grown 20x over the last four months, according to Huang, with a turning point in May when several strong open-weight models came out. Much of that inference runs on CoreWeave’s serverless platform. Cline is also moving beyond the integrated development environment and command line with Cline Desktop, an open-source app for open-weight models.
“Now with the agents getting better, with the model getting better, we are seeing this penetration trend into more of the general knowledge work, into people who are less sophisticated in coding tooling,” Huang said. “We introduce more services that can be more helpful to them.”
Open weights also change the math on long-tail projects, which rarely justify a big software budget or a dedicated engineering team. Control over both the model and the AI coding harness lets companies build custom tools for those niche jobs instead, Huang noted.
“You don’t really want to pay a huge standardized [software-as-a-service] budget if it’s a long-tail problem … [or] hire tons of engineers,” he said. “But now, because of how flexible things are, you can just build it.”
Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of Fully Connected:
(* Disclosure: TheCUBE is a paid media partner for the Fully Connected event. Neither CoreWeave, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)