> ## Documentation Index
> Fetch the complete documentation index at: https://docs.adcontextprotocol.org/llms.txt
> Use this file to discover all available pages before exploring further.

# AdCP and OpenRTB

> AdCP vs OpenRTB: how they differ and work together. AdCP handles agent-level campaign workflows; OpenRTB handles impression-level bidding. Complementary standards, not competing.

AdCP and OpenRTB are complementary standards that operate at different layers of the advertising stack. They are not competing — a platform can (and often will) implement both.

## What each standard does

|                    | OpenRTB                        | AdCP                                       |
| ------------------ | ------------------------------ | ------------------------------------------ |
| **Layer**          | Impression-level transactions  | Agent-level workflows                      |
| **Core operation** | Real-time bid request/response | Task-based campaign management             |
| **Participants**   | DSPs and SSPs                  | AI agents and advertising platforms        |
| **Timing**         | Real-time (milliseconds)       | Asynchronous (seconds to days)             |
| **Scope**          | Single impression auction      | End-to-end campaign lifecycle              |
| **Maintained by**  | IAB Tech Lab                   | AgenticAdvertising.org                     |
| **Maturity**       | Production (v2.6)              | Release candidate (v3.0-rc)                |
| **Transport**      | HTTP POST                      | MCP (tool calling) or A2A (agent-to-agent) |

## Where they overlap

Both standards touch media buying, but at different granularities:

* **OpenRTB** handles individual impression decisions: "Should I bid on this impression, and how much?"
* **AdCP** handles campaign-level decisions: "What inventory is available? Execute this campaign with this budget and targeting."

A single AdCP `create_media_buy` task might result in thousands of OpenRTB bid requests over the campaign's lifetime.

## Where they're different

**Scope.** OpenRTB is focused on the auction — bid requests, bid responses, win notifications, and billing events. AdCP covers the full campaign lifecycle: product discovery, creative management, audience activation, campaign execution, and delivery reporting.

**Communication model.** OpenRTB uses synchronous HTTP: a bid request arrives, and the bidder must respond within a few hundred milliseconds. AdCP is asynchronous: an agent submits a `create_media_buy` task, and the platform processes it on its own timeline, returning status updates.

**Participants.** OpenRTB connects demand-side platforms (DSPs) to supply-side platforms (SSPs) in automated auctions. AdCP connects AI agents to any advertising platform — including but not limited to DSPs and SSPs.

**Data model.** OpenRTB defines impression objects, bid objects, and deal objects. AdCP defines media products, media buys, creative formats, audience signals, and brand governance rules.

## How they work together

A typical integration uses both standards at different layers:

1. A **buyer agent** uses AdCP to discover available products on a publisher's platform (`get_products`)
2. The agent creates a campaign via AdCP (`create_media_buy`) with budget, targeting, and scheduling
3. The **publisher's ad server** executes the campaign by sending OpenRTB bid requests to the buyer's DSP for individual impressions
4. The buyer agent checks delivery via AdCP (`get_media_buy_delivery`) to monitor overall campaign performance

In this model, AdCP handles the strategic layer (what to buy, how much to spend, who to target) and OpenRTB handles the tactical layer (which specific impressions to win).

## Other standards in the ecosystem

AdCP and OpenRTB exist alongside several other standards:

| Standard                          | Purpose                                                       | Maintained by |
| --------------------------------- | ------------------------------------------------------------- | ------------- |
| **MCP** (Model Context Protocol)  | AI tool calling — how an AI model calls external tools        | Anthropic     |
| **A2A** (Agent-to-Agent Protocol) | Multi-agent collaboration — how autonomous agents communicate | Google        |
| **VAST** / **VPAID**              | Video ad serving and interactive video                        | IAB Tech Lab  |
| **ads.txt** / **sellers.json**    | Supply chain transparency and authorized seller verification  | IAB Tech Lab  |
| **Open Measurement SDK**          | Viewability and attention measurement                         | IAB Tech Lab  |

AdCP uses MCP and A2A as transport layers. It references IAB content taxonomies and audience segment standards where applicable.

## Frequently asked questions

<AccordionGroup>
  <Accordion title="Does AdCP replace OpenRTB?">
    No. They serve different purposes. OpenRTB handles real-time impression auctions. AdCP handles campaign-level agent workflows. A platform can implement both.
  </Accordion>

  <Accordion title="Do I need to implement OpenRTB to use AdCP?">
    No. AdCP works independently. A platform that doesn't use real-time bidding (for example, a direct-sold publisher or a commerce media network) can implement AdCP without any OpenRTB integration.
  </Accordion>

  <Accordion title="Can an AdCP agent trigger OpenRTB transactions?">
    Yes. When a buyer agent creates a media buy via AdCP, the sell-side platform can use any internal mechanism to fulfill the order — including OpenRTB auctions, direct insertion orders, or private marketplace deals.
  </Accordion>

  <Accordion title="Is AgenticAdvertising.org part of IAB Tech Lab?">
    No. AgenticAdvertising.org is an independent member organization. It is not a subsidiary, working group, or affiliate of IAB Tech Lab. However, AdCP is designed to be compatible with IAB standards.
  </Accordion>
</AccordionGroup>