Back to vendors
C

Cloudflare

Also known as: Cloudflare Agents, Cloudflare Agents SDK, Cloudflare Sandbox, Cloudflare Sandboxes, Cloudflare Sandbox SDK, cloudflare.com

Visit site
Entry priceWorkers Paid at $5 a month, with container usage included and per second rates above it.Full pricing detail

Cloudflare runs AI agents as Durable Objects with state, scheduling, memory, approvals, tracing and tools, alongside Workflows, AI Search, Browser Run and sandboxes for agent written code, on its global network.

Cloudflare runs a global network and developer platform, and its agent stack sits on that platform. The Agents SDK runs each agent as a Durable Object with its own SQLite storage, WebSockets, scheduling and durable jobs. Around it sit Workflows for durable multistep work, Workers AI and AI Gateway for models, AI Search and Vectorize for retrieval, Browser Run for driving Chrome, and sandboxes for agent written code in Containers, which are microVMs with their own kernel, or in Dynamic Workers for JavaScript, Python and WebAssembly.

Agents can be built on ready harnesses (Think, an opinionated chat agent framework, or Pi and OpenCode running inside a Durable Object) or on the lower level Lifecycle. The Session API keeps a branching conversation history with search and compaction, plus writable memory blocks the agent updates itself.

The MCP client connects to remote MCP servers with OAuth and elicitation, Code Mode turns MCP servers and OpenAPI services into typed connectors that model written code calls, and agents answer email, Slack, voice and webhooks.

Workflows can wait for human approval for months, durable Code Mode pauses chosen tool calls until someone approves them, and the Agents tab traces every turn with session replay and a waterfall of model calls, tool runs and approvals.

Agents bill as Workers, Durable Objects and model usage, and sandboxes need the $5 a month Workers Paid plan, which includes 375 vCPU minutes, 25 GiB hours of memory and 200 GB hours of disk for containers each month before per second rates apply. Cloudflare provides SOC 2, ISO and PCI reports to account administrators, operates FedRAMP Moderate data centers, and controls access through scoped roles, dashboard single sign on, SCIM and audit logs. Durable Objects can be confined to the EU, US or FedRAMP jurisdiction so agent state is stored and processed only there.

Cloudflare fits engineering teams building agents in TypeScript that want state, scheduling, approvals, retrieval, browsing and code execution in one runtime with a free hand over models. It fits less well for buyers who want a no code agent builder or a single product with its own price list, since the agent stack is assembled from several Cloudflare products, its Session memory APIs are still marked experimental, and Browser Run's Live View and session recording are in beta.

Vendor details

Canonical URL

https://www.cloudflare.com

Category

Agent infrastructure

Funding status

Cloudflare, Inc. is an independent, publicly listed company headquartered in San Francisco.

Company status

independent

Use cases & customers

Primary use cases

stateful agentsagent sandboxesbrowser agentsdurable workflowsMCP serverschat and voice agents

Target customers

developersplatform teamsAI product teams

Deployment options

SaaSAPI

Integrations

Agents connect to remote MCP servers with OAuth, to OpenAPI services and MCP tools through Code Mode connectors, to email, Slack, voice, webhooks and Chat SDK messengers such as Telegram, and to Workers AI, AI Gateway catalog models and vendor SDKs from the AI SDK, TanStack AI or Pi AI. Harnesses include Think, Pi and OpenCode, retrieval runs on AI Search and Vectorize, browsing on Browser Run, payments over x402 and the Machine Payments Protocol, and the platform is driven by the Cloudflare REST API, Wrangler and SDKs.

In practice

Your support agent has to remember each customer and follow up days later. On Cloudflare each customer gets an agent instance with its own storage, memory blocks it updates itself, and a scheduled callback that wakes it when the follow up is due.

Your agent can draft a refund but finance must approve it first. A Cloudflare Workflow holds the refund step with waitForApproval for as long as approval takes, then resumes the moment someone approves it from your app.

Your EU customers' agent data must stay in the EU. Create the agents' Durable Objects in the EU jurisdiction and their state is stored and processed only there, while the agents still serve users worldwide.

Agentic Index coverage score

13.0 / 14 capabilities · 93%

Integrations & Tool Calling Full

Agents act in outside systems through several routes. The MCP client connects an agent to remote MCP servers with OAuth, elicitation and reconnection after restarts, and Code Mode turns MCP servers and OpenAPI services into typed connectors that model written code calls while authentication stays in the host Worker. Agents also send and receive email, answer Slack and voice, receive webhooks, and can pay for HTTP services or MCP tools over x402 and the Machine Payments Protocol. Custom tools are ordinary functions handed to the model through the AI SDK, TanStack AI or a harness.

Sourcedevelopers.cloudflare.com/agents/tools/mcpread 2026-11-05

Workflow Orchestration Full

Multistep and multi agent work has its own runtimes. Workflows run durable steps with automatic retries and can wait months for an event or an approval, and the Agents SDK comes with patterns for prompt chaining, routing, parallelization and orchestrator and worker setups. Sub agents run as child Durable Objects with isolated storage and typed RPC, or as streaming tools of a parent agent, and a durable job queue and fibers keep long work alive through eviction and crashes.

Sourcedevelopers.cloudflare.com/agents/concepts/workflowsread 2026-11-05

Knowledge Grounding & RAG Full

Retrieval runs on maintained indexes. AI Search builds and manages indexes over R2 buckets, websites and uploaded files and returns relevant chunks, or a generated answer, to the agent, and Vectorize stores embeddings for custom retrieval. A searchable context block also gives each agent a full text index in its own Durable Object storage that it can query and add entries to.

Sourcedevelopers.cloudflare.com/agents/tools/ai-searchread 2026-11-05

Human Oversight & Guardrails Full

Approvals are part of the runtime. A Workflow can wait with waitForApproval() for minutes or months and resumes when the agent's approve or reject method is called, and durable Code Mode marks chosen connector calls with requiresApproval, so model written code pauses before the tool runs and replays once someone approves. MCP elicitation asks the user for input or an out of band step in the middle of a call, and a browser run can pause while a person logs in or solves a captcha through a Live View link.

Sourcedevelopers.cloudflare.com/agents/concepts/agentic-patterns/human-in-the-loopread 2026-11-05

Security, Identity & Governance Full

Cloudflare provides SOC 2, ISO and PCI reports to an account's Super Administrators and operates FedRAMP Moderate data centers. Access to the developer platform runs through account, domain and resource scoped roles, including Workers Platform Admin and its read only counterpart, Audit Logs Viewer and the Secrets Store roles, with dashboard single sign on, SCIM provisioning and audit logs. Agents check callers in the Worker before routing a request to them, and MCP servers built on Cloudflare can require OAuth 2.1 with scopes.

Sourcedevelopers.cloudflare.com/fundamentals/manage-members/rolesread 2026-11-05

Observability & Auditability Full

Agent activity is traced turn by turn. The Agents tab lists each agent's model, sessions, runs and token use, session replay shows messages, reasoning, tool calls and sub agent activity, and each trace is a waterfall of model calls, tool runs and approvals that exports through OpenTelemetry. Diagnostics channels emit structured events for schedules, jobs, workflows and MCP connections, and account audit logs record who changed what. Traces are kept for seven days, and recording message and tool payloads is opt in.

Sourcedevelopers.cloudflare.com/agents/runtime/observability/tracingread 2026-11-05

Memory & State Persistence Full

The Session API is a memory layer kept in each agent's own Durable Object SQLite database. Conversation history persists as a branching message tree with full text search and compaction that summarizes older messages without deleting them, and chosen messages or a whole session can be deleted.

Writable memory blocks, such as learned facts or a task list, are injected into every turn and updated by the agent itself through a set_context tool within a token budget. Memory belongs to the agent instance, usually one per user or conversation, and survives hibernation and eviction. These memory APIs are still marked experimental.

Sourcedevelopers.cloudflare.com/agents/concepts/conversation-state-and-memoryread 2026-11-05

Deployment & Data Residency Full

Agent state can be pinned to a jurisdiction. Durable Objects created in the EU, US or FedRAMP jurisdiction run and store data only there, location hints place new objects near one of eleven regions from western North America to Oceania, and Regional Services control where requests are answered. Agents run on Cloudflare's global network, and the routing helper passes jurisdiction and location hints straight to the agent's Durable Object.

Sourcedevelopers.cloudflare.com/durable-objects/reference/data-locationread 2026-11-05

Prebuilt Agents / Templates / Packs Partial

Starting points come as harnesses, examples and skills. Think is an opinionated chat agent framework with built in tools, memory, messengers and scheduled tasks, Pi and OpenCode harnesses run inside a Durable Object, examples cover browser, chat, email, Slack and voice agents, and the agents starter template ships ready to deploy, alongside Agent Skills for coding agents. These are frameworks and examples a developer builds on, not finished agents a business adopts as they are.

Sourcedevelopers.cloudflare.com/agents/harnessesread 2026-11-05

Triggers & Channel Coverage Full

Work reaches an agent from schedules, messages and events. The scheduler runs callbacks after a delay, at a date, on a cron expression or on an interval, and Think adds timezone aware recurring tasks. Agents receive email, Slack messages and mentions, voice over WebSocket, Chat SDK messenger webhooks such as Telegram, and webhook events routed to a dedicated agent instance, while a built in queue works through background tasks in order.

Sourcedevelopers.cloudflare.com/agents/agent-class/schedule-tasksread 2026-11-05

Model Flexibility & Routing Full

The developer chooses the model for each call. One createAI provider reaches Workers AI models hosted on Cloudflare, AI Gateway catalog models, dynamic routes and vendor SDK models from the AI SDK or Pi AI, with AI Gateway in front of outside providers, and harnesses such as OpenCode and Pi accept the same choices.

Sourcedevelopers.cloudflare.com/agents/modelsread 2026-11-05

APIs / SDKs / MCP Extensibility Full

Developers drive the platform through the Cloudflare REST API, Wrangler and SDKs, and each agent exposes its own methods to browsers and servers through callable methods, a typed client SDK and HTTP or WebSocket routes. Cloudflare's managed remote MCP servers let an assistant read configuration and manage services across an account, and the Agents SDK builds stateless remote MCP servers on Workers.

Sourcedevelopers.cloudflare.com/agents/agent-class/callable-methodsread 2026-11-05

Testing, Debugging & Optimization Partial

Agents can be tested before release. The Workers Vitest integration runs tests against an agent in the real runtime, checking its responses, methods and state, the agents starter template ships with it configured, and wrangler dev runs the agent locally. Nothing scores model output or compares an agent's behavior across versions.

Sourcedevelopers.cloudflare.com/agents/getting-started/testing-your-agentread 2026-11-05

Browser / Computer-use Full

Agents drive a real Chrome through Browser Run, which runs on the Workers Free and Paid plans. Model written code sends any Chrome DevTools Protocol command to a live session to navigate, run JavaScript, read the DOM and capture screenshots, and the agent can return to the same browser on later turns with its tabs, cookies and login intact while the session stays alive, up to 10 minutes idle by default. Quick actions return a page as Markdown, extracted data, its links or scraped elements. Live View and session recording, which let a person watch or replay a session, are in beta.

Sourcedevelopers.cloudflare.com/agents/tools/browserread 2026-11-05

The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded

Pricing

Workers Paid at $5 a month, with container usage included and per second rates above it.

Workers, Durable Objects and model usage, plus container vCPU, memory and disk per second

Included quota

Workers Paid includes 375 vCPU minutes, 25 GiB hours of memory and 200 GB hours of disk of container use a month, and 1 TB of container egress in North America and Europe (500 GB elsewhere).

What is public

Cloudflare publishes rates for each developer platform product, including the Workers Paid plan, container vCPU, memory, disk and egress, and container instance sizes.

Billing mechanics

Containers bill per 10ms of running time, with CPU on active use and memory and disk on the provisioned instance, and stop billing when an instance sleeps. Agents themselves bill as Worker requests and Durable Object usage, and model calls bill through Workers AI or the provider behind AI Gateway.

Cost watchouts

Container memory and disk bill on the provisioned instance size while the container runs, not on what it uses, and each container also carries its own Durable Object and Worker charges. Agent traces count toward account observability ingestion and storage from 1 December 2026.

Variable cost rationale

An agent's bill combines Worker requests, Durable Object compute and storage, model calls, browser time and container time, each metered separately, so busy agents and long running sandboxes can cost far more than the $5 plan.

Additional watchouts

A sandbox on the standard-3 instance (2 vCPU, 8 GiB) costs about $0.072 an hour in memory alone while it runs above the included allotment, plus $0.072 an hour per busy vCPU.

Overage / add-ons

Containers bill for every 10ms they run above the allotment, at $0.000020 per vCPU second of active CPU and $0.0000025 per GiB second of memory and $0.00000007 per GB second of disk on the provisioned instance size. Egress costs $0.025 per GB in North America and Europe, $0.05 in Oceania, Korea and Taiwan, and $0.04 elsewhere.

Sales call required

Mixed (some tiers require a call)

Free / trial

Sandboxes need the Workers Paid plan, which includes 375 vCPU minutes, 25 GiB hours of memory and 200 GB hours of disk for containers each month

Lowest paid plan

Workers Paid at $5 a month plus usage

Commercial notes

The agent stack runs on Cloudflare's developer platform, bought self serve on the Workers Paid plan or through an Enterprise agreement.

Key ambiguities

Agent costs are spread across Workers, Durable Objects, Workers AI or AI Gateway, AI Search, Browser Run and Containers price lists rather than one agent price.

Missing data

No single agent price exists; Enterprise terms are by agreement.

Agentic Index verified 2026-11-05

Contact us

Found a vendor we missed? Have feedback on the index? We'd love to hear from you.