WritingBasetenBasetenpublished Oct 22, 2025seen Jun 26

New Autoscaling Setting Target Utilization

Open original ↗

Captured source

source ↗
published Oct 22, 2025seen Jun 26captured Jun 27http 200method plain

New autoscaling setting: Target utilization Announcing our Series F . Learn more

changelog / post

New autoscaling setting: Target utilization

Oct 22, 2025 Go back

We've introduced a new autoscaling setting: Target utilization. This can be used to configure the amount of headroom you'd like on your model or Chain. Target utilization can be configured either via the UI or via our REST API . See our docs for more information. ✕ Target utilization uses your desired compute usage level to scale model replicas up or down.

Explore Baseten today Start deploying Talk to an engineer

Popular models GLM 5.2

Kimi K2.7 Code

DeepSeek V4

GPT OSS 120B

Whisper Large V3

NVIDIA Nemotron 3 Ultra

Explore all

Popular models GLM 5.2

Kimi K2.7 Code

DeepSeek V4

GPT OSS 120B

Whisper Large V3

NVIDIA Nemotron 3 Ultra

Explore all

Notability

notability 5.0/10

Autoscaling feature update, solid but not major.