Partnerships
Baseten Launches Public Preview of High-Performance Inference Inside Microsoft Foundry
By News Desk ·
Key takeaways
- Baseten on Microsoft Foundry is in public preview, with general availability expected in early 2027.
- The service offers DeepSeek V4.1 Flash 1017, GLM 5.3, GLM 5.3 Flash and Kimi K3 on Baseten's inference stack.
- Customers are billed through Azure as first-party Foundry consumption and can apply their existing Microsoft Azure Consumption Commitment.
- Baseten closed a $1.5 billion Series F in June 2026 at a valuation of up to $13 billion.
Baseten announced a partnership with Microsoft on November 17, 2026 that brings its high-performance inference to Microsoft Foundry, now in public preview. The GlobeNewswire release said general availability is expected in early 2027.
How Baseten on Microsoft Foundry Works
Enterprises on Azure can run frontier open models on Baseten's inference stack with Foundry's built-in security and governance, billed through their existing Microsoft agreement. The model lineup includes DeepSeek V4.1 Flash 1017, GLM 5.3, GLM 5.3 Flash and Kimi K3.
For residency requirements, Baseten offers global and U.S. data zones, along with zero data retention and Azure Policy controls that let administrators restrict specific models. Customers are billed through Azure as first-party Foundry consumption and can apply their existing Microsoft Azure Consumption Commitment. Teams can choose pay-as-you-go access or Provisioned Throughput Units for SLA-backed performance.
After an enterprise opts in, developers find Baseten models in Microsoft Foundry's model catalog, choose a model and select a deployment type. Deployments are ready in seconds, and no prompts or responses are stored.
"The next wave of AI innovation will come from what developers build when powerful open models meet high-performance infrastructure," said Steve Sweetman, VP of Product at Microsoft Foundry. "Bringing Baseten to Microsoft Foundry gives enterprises more choice and pairs Baseten's optimized and reliable inference with the governance, controls, and observability within the Foundry platform."
At Microsoft's Ignite event, Microsoft announced AI improvements focused on cheaper specialized models for specific tasks, including new models for coding, cybersecurity and decision-making, and Copilot Fusion to orchestrate multiple models at lower costs, according to Yahoo.
Baseten's Funding Before the Foundry Preview
According to akta.pro company records, Baseten closed a $1.5 billion Series F in June 2026 at a valuation of up to $13 billion across two tranches of $11 billion and $13 billion, its fourth fundraise in 18 months. The round was led by Altimeter Capital, Conviction and Spark Capital, with Sands Capital and Wellington Management as co-leads.
In January 2026, Baseten closed a $300 million Series E at a $5 billion valuation led by IVP and CapitalG, including a $150 million strategic investment from NVIDIA. The same records list nine funding rounds on file and total funding of $2.4 billion.
Baseten's Recent Product Rollouts
Baseten opened its Voice Agent Stack and Realtime Voice in private preview on November 16, 2026, with a waitlist, Unite.AI reported. The stack co-locates speech models and orchestration, reporting median turn latency of about one second across up to 56 concurrent sessions.
On November 6, 2026, Baseten said Serverless Loops was generally available, offering a managed SDK for supervised fine-tuning and reinforcement learning on open models, according to Unite.AI. Baseten joined NVIDIA's Nemotron Coalition on November 5, 2026, contributing enterprise inference expertise, Unite.AI reported.
Baseten's Early 2027 Target
Baseten said general availability of Baseten on Microsoft Foundry is expected in early 2027.