Anthropic’s Claude platform ranks third in enterprise agent stacks, trailing Microsoft and Salesforce

Anthropic’s Claude platform ranks third in enterprise agent stacks, trailing Microsoft and Salesforce

Claude captures 18% of enterprise agent deployments but struggles to become the primary platform companies build on

The enterprise AI agent market has its pecking order, and Anthropic isn’t at the top. Microsoft Copilot Studio commands 31% of enterprise agent deployments, Salesforce Agentforce holds 24%, and Anthropic’s Claude API sits at 18%, according to recent tracking data. Together, those three platforms account for 73% of all tracked enterprise agent deployments.

The primary platform problem

There’s a meaningful difference between a platform enterprises use and a platform enterprises build on. A February 2026 tracker measuring primary-platform adoption for agent orchestration painted a starker picture: Microsoft Copilot Studio led at 38.6%, OpenAI claimed 25.7%, and Anthropic trailed at 5.7%.

Microsoft and Salesforce enjoy deep integration into the software stacks enterprises already run. When your agent platform lives inside the same ecosystem as your CRM, your email, your cloud infrastructure, and your data warehouse, the switching costs become a moat. OpenAI’s Assistants and Responses API has achieved notably stronger primary-platform adoption within surveyed companies, maintaining a competitive gap that Anthropic has yet to close.

Advertisement

Where Anthropic actually wins

A late-2025 report found that Anthropic captured approximately 40% of enterprise LLM spend, overshadowing OpenAI’s 27%.

Anthropic launched Claude Managed Agents in 2026, expanding its portfolio beyond raw model access toward more structured deployment options. Revenue growth has been robust, tied directly to increased platform usage across enterprise customers.

A $7.6 billion market with room to run

The agent platform market was valued at $7.6 billion in 2025 and is projected to expand at a 43.84% compound annual growth rate through 2034.

Microsoft Copilot Studio’s 38.6% primary-platform adoption rate reflects its control of both the orchestration layer and the cloud infrastructure underneath it. Salesforce Agentforce benefits from sitting directly inside the CRM and customer data environments that drive revenue-generating workflows.

Anthropic launched Claude Managed Agents as one step toward closing the orchestration gap. Governance and safety features, areas where Anthropic has historically differentiated itself, could become a stronger selling point as regulatory frameworks around enterprise AI tighten.

The question facing Anthropic isn’t whether enterprises want to use Claude. They clearly do, given that 40% LLM spend figure. The question is whether Anthropic can translate model preference into platform commitment before Microsoft and Salesforce cement their positions as the default operating systems for enterprise AI agents.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
Anthropic’s Claude platform ranks third in enterprise agent stacks, trailing Microsoft and Salesforce
Anthropic’s Claude platform ranks third in enterprise agent stacks, trailing Microsoft and Salesforce

Claude captures 18% of enterprise agent deployments but struggles to become the primary platform companies build on

The enterprise AI agent market has its pecking order, and Anthropic isn’t at the top. Microsoft Copilot Studio commands 31% of enterprise agent deployments, Salesforce Agentforce holds 24%, and Anthropic’s Claude API sits at 18%, according to recent tracking data. Together, those three platforms account for 73% of all tracked enterprise agent deployments.

The primary platform problem

There’s a meaningful difference between a platform enterprises use and a platform enterprises build on. A February 2026 tracker measuring primary-platform adoption for agent orchestration painted a starker picture: Microsoft Copilot Studio led at 38.6%, OpenAI claimed 25.7%, and Anthropic trailed at 5.7%.

Microsoft and Salesforce enjoy deep integration into the software stacks enterprises already run. When your agent platform lives inside the same ecosystem as your CRM, your email, your cloud infrastructure, and your data warehouse, the switching costs become a moat. OpenAI’s Assistants and Responses API has achieved notably stronger primary-platform adoption within surveyed companies, maintaining a competitive gap that Anthropic has yet to close.

Advertisement

Where Anthropic actually wins

A late-2025 report found that Anthropic captured approximately 40% of enterprise LLM spend, overshadowing OpenAI’s 27%.

Anthropic launched Claude Managed Agents in 2026, expanding its portfolio beyond raw model access toward more structured deployment options. Revenue growth has been robust, tied directly to increased platform usage across enterprise customers.

A $7.6 billion market with room to run

The agent platform market was valued at $7.6 billion in 2025 and is projected to expand at a 43.84% compound annual growth rate through 2034.

Microsoft Copilot Studio’s 38.6% primary-platform adoption rate reflects its control of both the orchestration layer and the cloud infrastructure underneath it. Salesforce Agentforce benefits from sitting directly inside the CRM and customer data environments that drive revenue-generating workflows.

Anthropic launched Claude Managed Agents as one step toward closing the orchestration gap. Governance and safety features, areas where Anthropic has historically differentiated itself, could become a stronger selling point as regulatory frameworks around enterprise AI tighten.

The question facing Anthropic isn’t whether enterprises want to use Claude. They clearly do, given that 40% LLM spend figure. The question is whether Anthropic can translate model preference into platform commitment before Microsoft and Salesforce cement their positions as the default operating systems for enterprise AI agents.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.