OpenAI Presence: корпоративная платформа голосовых и чат-агентов для бизнеса

In July 2026, OpenAI unveiled Presence, an enterprise platform for deploying voice and chat agents that handle companies' customer-facing and internal workflows. OpenAI itself calls the platform "proven," but the announcement does not provide specific metrics, pricing, or a general availability date.

What OpenAI Presence is

OpenAI Presence is an enterprise platform for running AI agents in two formats: voice and text (chat). According to OpenAI's description, the platform helps organizations deploy "trusted" agents that work both in the external loop — with customers — and in the internal loop — with company employees.

• Name — OpenAI Presence
• Type — enterprise AI agent platform
• Two interaction formats: voice and chat agents
• Two use-case loops: customer service and internal processes
• Positioning — "proven," a solution tested in production

"A proven enterprise AI agent platform that helps organizations deploy trusted voice and chat agents for customer-facing and internal tasks," is how

OpenAI itself describes Presence in the announcement.

Why this matters for business

OpenAI Presence addresses the two most common enterprise AI agent scenarios at once: talking to customers by voice and corresponding with them in chat. Instead of building a voice bot and a chat bot on separate stacks, a company gets a single platform for both modalities and both loops — external customer inquiries and internal employee requests, such as IT or HR support.

The bet on the word "trusted" is key for the enterprise segment: businesses care that the agent doesn't fabricate answers and doesn't violate regulations. It is trust, not model "intelligence," that OpenAI puts front and center in the platform's description.

Voice and chat in OpenAI's strategy

Presence is not OpenAI's first agentic product. The company is already developing its agent lineup: the Operator browser agent and tools for building agents via the API. Presence carries this line into the enterprise loop, focusing on two channels — voice and chat.

OpenAI already has the technology base for voice agents: the company previously opened up the Realtime API for real-time voice interaction. Presence packages these capabilities into a ready-made solution for organizations, rather than a set of "building blocks" that a developer would need to assemble into an agent themselves.

What OpenAI has not disclosed yet

The Presence announcement doesn't answer the main questions a buyer would have. OpenAI has not disclosed four key parameters: pricing and connection terms, the general availability (GA) date, quality metrics and benchmarks, and the names of the customers on which the platform has already been "proven."

For an enterprise product, this is a significant gap: enterprise platforms are usually accompanied by security data, an SLA, and implementation case studies. Until those appear, the word "proven" in the text is not backed by public figures, and Presence can only be judged by its stated positioning.

What this means

OpenAI is moving from selling models and APIs to selling ready-made, turnkey enterprise agents — entering a market where Google, Microsoft, Salesforce, and dozens of specialized voice and chat bot vendors are already active. Presence is OpenAI's bid for this segment; how strong a bid it is will become clear once the company discloses pricing, metrics, and its first customers.