ToolsCompaniesAbout

Model Vault

by Cohere@cohereToronto, Ontario, CanadaNo ratings yet

Dedicated, fully managed inference for Cohere models

Visit cohere.com
Tool type
Dedicated model inference platform
Made for
Enterprise AI developers
Runs on
Managed cloud APIs and dashboard
Dedicated inference instances managed by Cohere; confirm region requirements with the vendor.
Last checked
28 Oct 2026
3 product sources

What it does

Model Vault is a dedicated, fully managed inference platform for running Cohere models. You deploy a model through the Cohere dashboard, and Cohere operates the serving infrastructure, including scaling, maintenance, updates, and GPU optimization. Applications send requests to the deployed model through an API-style interface, and the dashboard shows real-time request rates, latency, and token usage.

Each deployment is fully isolated, with no shared resources or multi-tenancy, and Cohere states that Confidential Computing keeps in-use data encrypted inside the hardware. A Model Vault can be created in minutes and gives access to Cohere embedding, reranker, and generative models. It can also serve as the inference layer for Cohere North, which Cohere says operates under a Zero Data Retention policy.

Model Vault suits enterprises and developers who want dedicated Cohere model inference without managing GPUs. Access requires a Cohere account and dashboard setup. Pricing is per instance by model and performance tier, so capacity is provisioned per deployment. Request a quote for custom enterprise needs, as well as for hosting regions, trial availability, and commitment terms.

What it can do

  • Dedicated, fully isolated model inference with no shared resources or multi-tenancy
  • Cohere-managed deployment, scaling, maintenance, and updates for Cohere models
  • Self-serve creation of a Model Vault in minutes through the Cohere dashboard
  • Access to Cohere embedding, reranker, and generative models
  • Real-time monitoring of request rates, latency, and token usage
  • Confidential Computing claimed to keep in-use data encrypted inside hardware

Works with

  • Cohere North

Example workflows

Deploy a model for production inference
Input: a selected Cohere model and performance tier. Steps: create a Model Vault in the Cohere dashboard, choose a billing basis (hourly or monthly/annual commitment), send API requests to the deployed model, and monitor request rate, latency, and token usage. Output: a dedicated inference endpoint with usage metrics.
Use Model Vault as the inference layer for North
Input: a North deployment running on CPU-only infrastructure. Steps: point North's inference calls to a Model Vault instance, let Cohere handle GPU optimization and scaling, and monitor throughput in the dashboard. Output: North application logic runs without GPU provisioning, with inference served by the isolated Model Vault.

Before you choose

  • Pricing is per instance by model and performance tier; request a quote for custom enterprise needs.
  • Dedicated instances are not shared, so capacity is provisioned per deployment rather than billed purely per token.
  • Confidential Computing and Zero Data Retention are vendor claims; verify them against your compliance requirements.

Listed under

Get started

Model Vault is set up through the Cohere dashboard, or via sales contact.

  1. Sign in to the Cohere dashboard at dashboard.cohere.com and create a Model Vault.
  2. Select the model and performance tier, then choose hourly or longer-term billing.
  3. Connect your application or North deployment to the Model Vault endpoint and monitor usage in the dashboard.
Sign in to the Cohere dashboard or contact sales

Plans and prices

Prices in USD

Model Vault is billed per instance based on model and performance tier, hourly or via monthly/annual commitments. Free trial API keys are described for general API usage, but whether they apply to Model Vault is not stated. Contact sales for custom enterprise pricing.

Embed 5 Fast, Small tier$3 per instance hourHourly rate per instance; monthly rate listed separately at $2,000.
  • Fully managed model deployment
  • No shared resources
Rerank 5 Pro, XL tier$7 per instance hourHourly rate per instance; monthly rate listed separately at $4,300. Tier mapping in the extracted table is ambiguous.
  • Fully managed model deployment
  • Simple startup and self-serve model access

Source: cohere.com/pricing, checked 28 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 28 Oct 2026. Each source supports part of the overview; availability and prices can change.

Ratings and reviews

No ratings for Model Vault yet.

Be the first to rate it

What’s changed

Free account

See what’s changed at Model Vault

Launches, price moves and pivots, each with a date and a source. Free with an account.

Signals

Free account

See Model Vault’s signals

Launch traction, hiring and release activity, tracked over time. Free with an account.

Data and security

Free account

See how Model Vault handles your data

Training on your files, compliance, hosting and the models behind it. Free with an account.

Compare with similar tools

Free account

Compare Model Vault side by side

Prices, platforms and features against similar tools, with sources. Free with an account.

Where this comes from

How we research
Company website
cohere.com: product, pricing and policy pages
YC directory mirror
Batch and company name
Web search
Confirms the official site
Social accounts
Accounts the company links, and its posts on X
Human review
Checked by a person
knowledge_base

Last checked 28 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.