OpenAI is ahead of Anthropic in one important part of the enterprise artificial intelligence market: the software businesses use to build, deploy and manage AI agents. A September 2026 survey by VentureBeat Intelligence found that 43% of companies using an AI agent-orchestration platform reported using OpenAI’s tools, compared with 22% using Anthropic’s platform. Google’s platform was used by 21% of respondents in the same group.
The findings highlight an increasingly important distinction in the AI industry. Companies are not simply choosing which model generates the best answers; they are also choosing the infrastructure that connects models to business applications, authorizes tool use and manages agent workflows. OpenAI’s lead in reported platform adoption gives it an advantage in this layer of the market, although the survey does not establish that it leads Anthropic in every measure of enterprise AI usage, spending or model preference.
Key Takeaways
- OpenAI’s agent-orchestration tools were used by 43% of surveyed companies using such platforms, compared with 22% for Anthropic and 21% for Google.
- Among 107 companies using an agent-orchestration platform, 25% named OpenAI as their primary platform, ahead of Google’s 18% and Anthropic’s 16%.
- OpenAI’s products in the survey were the Agents SDK and Responses API; Anthropic’s were Claude Agent SDK and Managed Agents.
- Among companies considering a new, additional or replacement platform, Anthropic appeared on 33% of respondents’ shortlists, compared with 30% for OpenAI.
- The results come from a self-selected survey of enterprise-related respondents, so they should not be treated as a census of the entire business market.
- The findings show that platform adoption, model usage and corporate spending are different measures of competition.
What the VentureBeat Survey Found
VentureBeat Intelligence conducted its agent-orchestration survey in September 2026. The survey included 109 qualified respondents, while 107 companies using an agent-orchestration platform formed the base for the platform-usage comparison.
Respondents could identify multiple platforms their organizations used. This means the percentages are not mutually exclusive market shares: a company using OpenAI and Anthropic could count toward both figures.
| AI platform provider | Share of platform users reporting usage |
|---|---|
| OpenAI | 43% |
| Anthropic | 22% |
| 21% |
Source: VentureBeat Intelligence, September 2026 survey. Base: 107 companies using an agent-orchestration platform. Multiple responses were permitted.
OpenAI’s reported usage rate was almost twice Anthropic’s. However, the result measures whether surveyed companies use a platform, not how much they spend on it, how many agents they run or how deeply the platform is integrated into their operations.
The survey also asked companies to name their primary agent-orchestration platform. In this question, respondents selected the platform they relied on most rather than listing every platform in use.
| Primary platform | Share of respondents naming it primary |
|---|---|
| OpenAI | 25% |
| 18% | |
| Anthropic | 16% |
OpenAI also led on this measure, while Google placed ahead of Anthropic. The gap between the usage and primary-platform figures is important: installing or testing a platform does not necessarily mean a company has made it the foundation of its agent infrastructure.
What Is an AI Agent-Orchestration Platform?
An AI agent-orchestration platform is the software layer that helps an AI agent carry out tasks beyond generating a single response. It can connect models to tools, execute workflows, manage context and enforce permissions.
For example, a company might deploy an agent that reviews customer support requests, checks an order-management system, drafts a response and escalates unusual cases to a human employee. The underlying language model generates and reasons about content, while the orchestration layer manages the sequence of actions and the systems the agent can access.
The platform can therefore influence how quickly a business builds an application, how reliably the workflow operates and how much control administrators have over its execution.
In the survey, OpenAI’s offering included its Agents SDK and Responses API. Anthropic’s included Claude Agent SDK and Managed Agents, while Google’s offering was its Enterprise Agent Platform.
These tools compete for a position that can become difficult to replace once an organization has built applications, permissions, monitoring and internal processes around them.
Why OpenAI’s Lead Matters
OpenAI’s reported advantage suggests that its developer ecosystem is gaining traction among the companies represented in the survey. A platform that is already being used across an organization may have an advantage when teams choose where to build their next AI application.
Once developers have connected business systems, written agent instructions, configured tools and created evaluation procedures, moving to another platform can require substantial rework. That creates a practical incentive to expand an existing deployment rather than rebuild it elsewhere.
The advantage may become more valuable as businesses move from isolated AI experiments to workflows that perform multiple steps. A company deploying agents across customer support, software development, finance and internal operations may prefer a common platform for monitoring, permissions and execution.
However, reported adoption alone does not establish that OpenAI’s tools are technically superior. Businesses can choose a platform for many reasons, including ease of development, reliability, model compatibility, operational controls, pricing and the expertise of their existing teams.
The platform that attracts the most users is not automatically the best option for every company.
Anthropic Remains a Serious Competitor
Anthropic’s lower reported usage does not mean it has lost its opportunity in enterprise AI. The same September survey asked companies planning to adopt, add or replace an agent-orchestration platform within the next 12 months which products they were considering.
Among the 61 companies that named at least one platform under consideration, 33% included Anthropic, compared with 30% for OpenAI and 25% for Google. Respondents could select multiple platforms.
| Platform | Share of companies considering it |
|---|---|
| Anthropic | 33% |
| OpenAI | 30% |
| 25% |
Source: VentureBeat Intelligence, September 2026. Base: 61 companies naming at least one platform under consideration. Multiple responses were permitted.
This result provides a different perspective from the current-usage figures. Anthropic appeared on slightly more shortlists than OpenAI, suggesting that businesses already using other platforms may still be evaluating Anthropic for future projects.
The figures do not show how many companies will ultimately adopt each platform. Being considered is an early stage in the purchasing process, and some evaluations will not lead to deployment.
Still, the results indicate that Anthropic remains relevant to enterprise buyers despite its lower reported current usage in this survey.
Model Choice and Platform Choice Are Different
The survey also highlights why it can be misleading to judge the enterprise AI race using a single metric. A company may use OpenAI’s orchestration tools while selecting models from several providers. Another may run Claude models through a cloud provider or use Anthropic’s platform alongside a different primary system.
VentureBeat’s separate August survey found that 69% of companies using OpenAI’s Agents SDK or Responses API named OpenAI as their primary orchestration platform. For Anthropic’s Claude Platform and Agent Skills, the corresponding figure was 38%.
That earlier survey used a different sample and should not be directly combined with September’s figures as a continuous trend. It nevertheless reinforces the importance of distinguishing platform installation from primary-platform status.
VentureBeat’s October 5 report also found that only 2% of surveyed enterprise platform users said alignment with a leading AI model was the main factor influencing their platform choice. Flexibility across models and tools, production reliability, ease of development and control over agent execution were more prominent considerations.
This suggests that businesses may value a platform’s ability to manage the entire workflow more than exclusive access to one model.
Enterprise AI Adoption Still Has Measurement Challenges
The survey offers useful insight, but its methodology places limits on what can be concluded.
Respondents came from VentureBeat’s readership and a research panel. Companies qualified if respondents worked at organizations with at least 100 employees or consulted for such an organization as their main client. The sample was therefore self-selected rather than a probability sample representing all enterprises.
The results describe the companies represented in the survey. They cannot establish the exact share of all businesses worldwide using each provider.
The questions also measured different things: platform usage, the primary platform and future consideration. Those indicators should not be merged into one overall market-share figure.
Finally, a platform’s presence in a company does not reveal how extensively it is used. One organization might run a small pilot while another depends on the same platform for mission-critical operations. Both may answer that they use it, despite very different levels of adoption.
Future surveys with consistent samples, more detailed measures of deployment scale and information about spending could provide a clearer picture of the market.
The Bigger Picture
The competition between OpenAI and Anthropic is expanding beyond which model performs best on a benchmark. Both companies are developing the software needed to turn models into agents that can use tools, access information and execute multi-step business workflows. The orchestration layer matters because it determines how those capabilities are integrated into real operations.
OpenAI’s lead in VentureBeat’s September platform-usage survey is an encouraging sign for its enterprise developer ecosystem. Yet Anthropic’s presence on more respondents’ shortlists shows that current adoption does not necessarily predict future choices. Businesses are still evaluating their options, and the market remains competitive.
Looking Ahead
The next question is whether OpenAI can convert its reported usage advantage into deeper, more durable deployments. Continued growth will depend on reliability, ease of development, governance, pricing and how well its tools fit the systems companies already use. Anthropic, meanwhile, could gain ground if organizations evaluating its tools move from experimentation to production.
For enterprise buyers, the practical decision is not simply to follow the provider with the highest survey percentage. Companies should compare platforms against their own requirements, test integrations, measure operational costs and examine security and permission controls before committing to a long-term architecture. As AI agents become more widely deployed, the most successful platform may be the one that makes them dependable and manageable—not merely the one with the largest reported user base.
Frequently Asked Questions
1. How many companies use OpenAI’s agent platform compared with Anthropic’s?
In VentureBeat Intelligence’s September 2026 survey, 43% of companies using an agent-orchestration platform reported using OpenAI’s tools, compared with 22% for Anthropic. Respondents could select multiple platforms.
2. Which AI agent platforms were included?
The survey identified OpenAI’s Agents SDK and Responses API, Anthropic’s Claude Agent SDK and Managed Agents, and Google’s Enterprise Agent Platform, alongside other enterprise and developer tools.
3. Which provider was named the primary platform most often?
Among the surveyed companies naming a primary platform, OpenAI received 25% of responses, Google 18% and Anthropic 16%.
4. Is Anthropic losing the enterprise AI race?
The survey does not establish that. Anthropic had lower reported current usage in this sample but appeared on 33% of the shortlists from companies considering platforms, compared with 30% for OpenAI.
5. Does the survey represent every company?
No. The survey used a self-selected group of respondents associated with organizations of at least 100 employees. Its findings are a snapshot of those respondents, not a definitive measurement of all enterprises.
6. Why do businesses need agent-orchestration platforms?
These platforms help connect AI models to tools and business systems, manage multi-step tasks, enforce permissions and support monitoring. They make it easier to deploy and govern agents beyond a standalone chatbot interaction.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
