Pulumi AI

Pulumi AI is Pulumi’s AI-powered approach to cloud infrastructure management, now centered on Pulumi Neo, an AI infrastructure agent designed to help developers and platform teams provision, manage, troubleshoot, govern, and optimize cloud infrastructure. Pulumi’s current Infrastructure AI platform also allows teams to use their existing AI coding assistants, including Claude Code, Codex, Cursor, and GitHub Copilot, through Pulumi Agent Skills and the Pulumi MCP server.

Pulumi itself is an open-source infrastructure-as-code platform that allows developers to define cloud infrastructure using familiar programming languages rather than relying exclusively on configuration languages. Pulumi supports major cloud platforms including AWS, Microsoft Azure, Google Cloud, and Kubernetes.

The original Pulumi AI was introduced as a way to use natural-language prompts to discover cloud infrastructure solutions and generate Pulumi Infrastructure as Code. Pulumi later developed these capabilities into Pulumi Copilot and then Pulumi Neo. Pulumi’s official documentation now states that the previous Pulumi AI project-generation API has been retired and recommends Pulumi Neo instead.

Pulumi Neo

Pulumi Neo is designed as an AI platform engineer rather than simply a conversational coding assistant. It can investigate live infrastructure, understand Pulumi projects and stacks, review infrastructure code, propose changes, generate infrastructure code, and execute multi-step infrastructure tasks.

Users can describe infrastructure requirements in natural language, after which Neo can break the request into steps, create a plan, execute appropriate changes, and operate within Pulumi’s governance controls. It can also identify infrastructure issues, investigate deployment failures, and help optimize cloud environments.

Infrastructure Automation

Pulumi AI is particularly useful for infrastructure automation. Neo can perform complex, multi-step cloud operations rather than simply returning code suggestions. It can work with existing infrastructure, investigate dependencies, create proposed changes, and automate recurring maintenance tasks.

This makes it useful for platform engineering teams managing infrastructure across multiple environments and cloud providers.

AI-Assisted Infrastructure as Code

Pulumi allows AI coding agents to work directly with real infrastructure code. Because Pulumi programs produce a verifiable infrastructure plan, AI agents can generate code, preview proposed changes, and deploy infrastructure through Pulumi’s workflow.

The Pulumi MCP server extends this functionality to compatible AI assistants. It can provide AI assistants with access to Pulumi Cloud stacks and resources, the Pulumi Registry, policy reports, and infrastructure search. AI assistants can also delegate complex tasks to Pulumi Neo.

AI Governance and Security

A major strength of Pulumi AI is that AI-driven infrastructure changes can remain subject to organizational controls. Pulumi Neo supports RBAC, policy guardrails, human approval workflows, and audit trails. This is particularly important when AI agents are allowed to make changes to cloud infrastructure.

Neo also provides different operating modes that let teams choose how much autonomy the agent has. Review and Balanced modes allow greater human oversight, while Auto mode is intended for trusted workflows where Neo can operate with less intervention.

AI Infrastructure Insights

Pulumi AI can help teams investigate existing cloud environments rather than only creating new infrastructure. Users can query their Pulumi resources and cloud infrastructure, search for resources, investigate policy violations, identify problems, and obtain infrastructure-related insights through AI assistants connected to Pulumi.

Neo can also work with external tools through MCP integrations. Pulumi has documented integrations with services including Atlassian, Datadog, Honeycomb, Linear, PagerDuty, and Supabase, alongside cloud CLI access for AWS, Google Cloud, Azure, and Kubernetes.

Supported Development Languages

Pulumi is a multi-language infrastructure-as-code platform. Its currently documented languages include:

• TypeScript
• JavaScript
• Python
• Go
• C#
• Visual Basic
• F#
• Java
• Pulumi YAML
• HCL

Pulumi’s AI capabilities can generate or work with infrastructure code in Pulumi-supported environments, while the current AI SDKs and integrations have their own language-specific support.

Pricing

Pulumi currently uses a combination of free and usage-based pricing for its platform.

Individual — $0/month

• Free forever for individuals
• 1 user
• IaC state management
• Unlimited projects, stacks, and environments
• Up to 500 workflow minutes
• Pulumi Neo includes 5 million free tokens per month

Team — $40/month base

• Up to 10 users
• 40 Pulumi Credits included
• Additional usage billed according to consumption
• Neo-powered features are metered through Neo tokens
• Includes Neo code reviews and Slack integration
• Free trial available

Enterprise — $400/month base

• 400 Pulumi Credits included
• Up to 2,000 resources
• Additional usage billed on demand
• Enterprise security and collaboration capabilities
• Free trial available

Business Critical — Custom pricing

• Advanced governance, policies, and controls
• Self-hosting available
• Compliance capabilities
• SCIM
• Audit-log export
• Volume pricing
• Optional 24/7 enterprise support

Pulumi currently lists $3 per 1 million Neo tokens for additional Neo usage on paid plans. One Pulumi Credit is equivalent to $1 USD.

Pricing Model

Usage-based pricing combined with subscription plans. Pulumi Cloud uses Pulumi Credits for consumption, while Pulumi Neo uses token-based metering for AI-powered activities.

Pricing Note

The Individual plan is free forever and includes 5 million Neo tokens per month. Team and Enterprise plans include monthly Pulumi Credits, with additional usage billed according to consumption. Business Critical uses custom pricing. AI usage is therefore not simply priced as a conventional per-user AI subscription.

Review

Pulumi AI has evolved significantly from its original natural-language infrastructure generator. The current Pulumi Neo experience is much more focused on agentic infrastructure management, allowing AI to investigate, plan, generate, review, and execute infrastructure tasks while operating within Pulumi’s existing infrastructure-as-code and governance model.

Its strongest advantage is the combination of AI with real infrastructure workflows. Instead of producing isolated infrastructure suggestions, Neo can work with Pulumi stacks, resources, policies, and deployments. The MCP server also makes Pulumi’s infrastructure context available to AI coding assistants such as Cursor, Claude Code, Windsurf, and other MCP-compatible tools.

The platform is particularly attractive to DevOps, platform engineering, cloud engineering, and infrastructure teams that want to automate repetitive cloud operations while retaining governance and human oversight.

The main consideration is that “Pulumi AI” as a standalone product is no longer the current product experience. Pulumi has retired the earlier AI project-generation service and directs users toward Pulumi Neo.

Pulumi is therefore best viewed today as an infrastructure-as-code platform with an integrated AI layer, rather than as a general-purpose AI assistant.