Baseten is now available as a first-class AI provider in Runlayer. You can use open-source models like Kimi-K3 GLM-5.2, DeepSeek-V4 Pro running on Baseten's inference platform, with the same identity-aware permissions, observability, cost controls, and runtime security that govern every agent on Runlayer.
Select Baseten under Settings → AI Provider. Drop in your API key, and every model in your Baseten workspace is ready to power your team’s agents. Every model supports OpenAI Compatible and Anthropic Messages API formats.
You can then select which model you want to run per agent.
The right model for the right task
Many agent workloads require a balance of quality, performance and cost. Open models like GLM-5.2 and Kimi K3 deliver excellent results at a fraction of the cost. Our integration with Baseten allows you to use frontier open or custom models for each individual agent while Runlayer ensures governance and security aren’t an afterthought.
- Baseten gives you the inference layer: an inference platform with the tools to deploy frontier open-source models on infrastructure built for production, with the latency, throughput, and reliability your agents and enterprise demands.
- Runlayer gives you the enablement and control layer: connectors to your real tools and data, skills your whole team can use, and the visibility and control your security team requires.
Get started
Baseten is live in the AI provider settings for all Runlayer customers today. If you're already running models on Baseten, connecting takes about a minute.
