Flexible pricing that scales as you grow

Pay-as-you-go API pricing, enterprise custom deployments, and personal plans.

Context Engine API

API access to reasoning-ready outputs from email, work apps and the live web. Pay-as-you-go, no setup required.

CEF-1 Normal

Best for high-volume automation, summaries, and routine workflows.

PricingInput:$0.48 / 1M tokens
Output:$1.92 / 1M tokens
Get Started

CEF-1 High

Best for critical workflows where accuracy, depth, and long context matter.

PricingInput:$2.40 / 1M tokens
Output:$9.60 / 1M tokens
Get Started

Early Access: Unlimited evaluation credits for a limited period

Includes: $10 to test in the playground

See how your organization can work brilliantly with iGPT

Frequently Asked Questions

Have more questions?

iGPT for enterprise is an intelligence layer that connects your email, files, tools, and live web data, then turns that raw data into reasoning-ready answers with citations and structured outputs.

Search helps you retrieve, while iGPT reconstructs context. It identifies decisions, owners, risks, sentiment shifts, and next steps across threads and tools, then answers in a way that reflects how work actually unfolded.

iGPT mirrors existing permissions from your connected systems and enforces them end-to-end. Users only see what they are already allowed to access, with centralized admin controls, RBAC, policy enforcement, and audit visibility.

No model training on customer data. For inference, the platform is designed for zero data retention, alongside enterprise controls and auditability.

Three models: cloud, hybrid, and on-prem. Cloud is fully managed. Hybrid keeps data in your VPC, while inference runs in the cloud. On-prem runs the entire platform inside your infrastructure.

Yes. Teams can create custom agents with no code by selecting data sources, setting capabilities, and deploying internally. The same intelligence core powers chat, search, and agents across the organization.

iGPT connects to your email, calendar, documents, and tools through secure OAuth. When you ask a question, it searches your real conversations and files, rebuilds the context of each thread, understands who said what and when, and combines that with live web data. The answer is based on your actual work, not guesses.

Only what you already have permission to see. iGPT follows the same access rules as Gmail, Drive, Slack, and other connected tools. If you cannot view a file or email, iGPT cannot use it. Each person has a private AI instance that cannot access anyone else's data.

No. Your emails, documents, and chat history with iGPT are never used for model training. All data is encrypted in transit and at rest, and is used only to answer your questions. If you disconnect a data source or delete a file, it is removed from iGPT's index.

In the personal chat interface, your conversations are private. Admins cannot view your prompts or responses. In enterprise deployments, organizations can enable audit logs if required by compliance policies, but the default setting keeps individual usage private.

ChatGPT and Claude do not have access to your email threads or documents. They cannot track decisions, reconstruct conversations, or understand the context of your work. iGPT is designed specifically for business communication, linking your internal knowledge with real-time web data to give accurate and actionable answers.

You connect your tools through secure OAuth, the same way you would connect apps like Zapier. iGPT begins syncing your recent emails, files, and messages in the background. You can start asking questions immediately, and the sync continues automatically so your context stays current.