Standard Vault Overview
A Standard Vault is a Cohere-managed inference environment on dedicated infrastructure. You get full control over model selection, scaling, and performance monitoring, without operating the underlying serving stack. Your data is protected in transit and at rest, and Cohere manages maintenance, deployments, updates, and scaling.
A Standard Vault runs Cohereโs models alongside leading third-party open models, such as DeepSeek, GLM, MiMo, Gemma, Nemotron, GPT-OSS, Phi, and Qwen, in the region of your choice: the United States, the European Union, or Canada.
Choose Standard Vault when you want dedicated inference and donโt require hardware-backed protection of data in use or verifiable remote attestation. If you do, see Encrypted Vaults.
Deploying and using a Standard Vault
Standard vaults use the shared vault flow:
Whatโs specific to Standard Vault
Architecture
These are a Standard Vaultโs core architectural components:
- Logically isolated: Isolates all infrastructure components, including the network load balancer, reverse proxy, serving middleware, inference servers, and GPU accelerators.
- Minimal shared components: Shared infrastructure is limited to authentication and underlying Kubernetes/compute resources (nodes, CPU, and memory).
- Cohere-managed operations: Cohere handles maintenance, deployments, updates, and scaling.
Zero Data Retention (ZDR) is on by default in Model Vault: Cohere processes inputs and outputs for inference but does not retain any prompts or responses.