Arcjet vs DataDome

Arcjet vs DataDome: DataDome integrates at the edge with dashboard policy.

13 min read
In short: DataDome integrates at your CDN, web server, or backend, rates each AI agent's identification strength, and applies per-agent policies from its dashboard. Arcjet runs in your request handler and returns a bot identity with a kind, evidence level, and delegated flag, verifying both Web Bot Auth and Personal Agent Protocol sessions, so your code can combine it with the user, cart, or plan.

Arcjet vs DataDome

If you're looking for a DataDome alternative for developers, the main difference is where the decision lands. DataDome is a bot and fraud protection platform that integrates at your CDN, web server, or backend and recommends the edge, with policy managed in its dashboard. Arcjet is an AI agent runtime security platform whose bot detection runs in your request handler and returns a bot identity your code reads: what kind of bot it is, which bot, and how strongly Arcjet verified it.

Both products identify AI agents, separate them from AI crawlers, and verify Web Bot Auth signatures. DataDome rates each agent's identification strength and lets a security team set per-agent and per-endpoint policies in its dashboard. Arcjet reports the evidence on each decision, verifies Personal Agent Protocol (Poppy) sessions from vetted agents, and lets the handler combine the identity with the user, cart, or plan. Choose DataDome if a fraud or security team wants one dashboard across many CDNs, web servers, and mobile apps. Choose Arcjet if developers want bot decisions as code, with application context and transparent usage pricing.

Arcjet publishes this comparison. Competitor details come from DataDome's public documentation, reviewed on November 16, 2026. Products change, so check the linked sources before you decide.

Key differences: DataDome vs Arcjet

DataDome and Arcjet differ in six ways: deployment model, AI agent handling, evidence, signed and delegated agent support, developer workflow, and pricing transparency. The following sections compare each one.

Deployment model

DataDome integrates where your traffic passes, and Arcjet integrates in your code. DataDome's getting started guide lists integrations for CDNs and edge platforms (Akamai EdgeWorker, AWS CloudFront, Cloudflare Workers, and Fastly), web servers and gateways (Nginx, Apache, HAProxy, Envoy, and Kong), and backends (Node.js, Go, Python ASGI, Java, Ruby, and Next.js), and states "we strongly recommend integrating DataDome at the Edge (CDN level) whenever possible." A JavaScript tag and iOS and Android SDKs collect client signals.

A DataDome server module calls DataDome for each request. The Node.js module calls datadomeClient.handleRequest, with a default timeout of 150 ms, "after which the request will be allowed."

Arcjet runs as an SDK in your request handler, with SDKs for JavaScript and TypeScript, Python, and Go. The handler calls protect() and receives a decision that includes the bot identity. Arcjet doesn't sit in front of your origin, so it doesn't absorb network-layer DDoS attacks, and it works behind any CDN.

AI agent handling

Both products separate AI agents from AI crawlers, with different category schemes. DataDome's AI agent identification uses four categories: AI Crawler, AI Assistant, Autonomous Agent, and Agentic Browser. Its Access Control applies Intent-Based, Allow, Block, Rate Limit, Time-Box, or Monetize policies at three levels: a default per agent, a policy per endpoint, and custom rules. DataDome applies "out-of-the-box recommended policies for all known AI agents."

Arcjet sets a kind on each identity: AI_AGENT for a bot that fetches on demand for a waiting person, AI_CRAWLER for one that collects pages ahead of time, AUTOMATION for browser automation that advanced signals detected, or OTHER. A bot rule allows or denies by bot ID or category, such as CATEGORY:AI_AGENT allowed and CATEGORY:AI denied, and the helpers isAiAgent, is_ai_agent, and IsAIAgent() check for a live AI agent identity that isn't spoofed. For more information, see how to block AI crawlers but allow AI agents.

Evidence and identification strength

Both products grade how an agent was identified, and they expose the grade in different places. DataDome assigns each identified agent an identification strength. High-strength methods are Web Bot Auth, cryptographic token verification, trusted IP lists, and reverse DNS. Network attribution and client-side fingerprinting are medium, and server-side fingerprinting is low. DataDome shows the strength in its Access Control and Explore dashboards, and its trust score combines identification strength with fraud history.

DataDome's enriched headers give your server X-DataDome-botname, X-DataDome-isbot, and X-DataDome-ruletype, whose values include Verified Bots and AI Agents. A header for identification strength isn't documented. Advanced headers, such as a bot score and threat markers, need DataDome support to turn them on.

Arcjet reports evidence on each decision: UNVERIFIED, FINGERPRINT, IP, or SIGNATURE, from weakest to strongest. It sets spoofed when IP or reverse DNS verification contradicts a claim. Your handler can require SIGNATURE on checkout and accept IP on a product page in the same function. For more information, see AI agent evidence levels.

Signed and delegated agents

Both products verify Web Bot Auth, and only Arcjet documents verifying Poppy sessions, as of November 2026. DataDome's bot authentication docs state that it "can cryptographically verify http signatures using the Web Bot Auth protocol". They also list reverse DNS, static and dynamic IP lists, and private AS checks, and state that unauthenticated automated requests are categorized as Threat Detection and blocked by default. DataDome's documentation doesn't mention the Personal Agent Protocol.

Arcjet verifies Web Bot Auth for the signers in its open well-known-bots catalog and for any directory you name with SIGNED_BOT:. It verifies Poppy sessions for agents vetted in the same catalog. A verified session reports kind AI_AGENT, evidence SIGNATURE, the agent's domain as operator, and delegated true. Arcjet doesn't fetch a client_id that a request supplies. A Poppy session doesn't identify the user and doesn't prove the user approved a specific action, so your authorization server checks the access token and scopes. For more information, see the Personal Agent Protocol.

Developer workflow

DataDome is configured mostly in its dashboard, and Arcjet is configured mostly in code. DataDome's policies live in Access Control, and your server reads the module's result and enriched headers. Arcjet rules sit next to the route they protect, so they go through code review, run the same way locally as in production, and can run in dry run before you enforce them. Operational remote rules for bots, Shield, filters, and rate limits change in the Arcjet Console, CLI, or MCP server without a redeploy.

Pricing transparency

Arcjet publishes its prices, and DataDome's documentation describes plans by request volume. DataDome's Plan and Traffic page shows "your plan linked to a number of requests," and log enrichment requires a Corporate or Enterprise plan. Arcjet has Individual ($25 per month), Startup ($299 per month), and custom Enterprise plans, plus $5 per million web requests. For details, see Arcjet pricing.

Arcjet vs DataDome comparison table

The following table compares Arcjet and DataDome across 11 areas, from where each one runs to what it covers beyond bots.

AreaArcjetDataDome
Where it runsSDK in your request handler, on any host and behind any CDN.

CDN, web server, or backend integration, with the edge recommended, plus a JavaScript tag and mobile SDKs.

LanguagesJavaScript and TypeScript, Python, and Go.

Server modules for Node.js, Go, Python ASGI, Java, Ruby, ASP.NET Core, and others, plus CDN and web server integrations.

AI agents vs AI crawlers

kind: AI_AGENT, AI_CRAWLER, AUTOMATION, or OTHER.

AI Crawler, AI Assistant, Autonomous Agent, and Agentic Browser.
Web Bot Auth

Verified for catalog signers and any directory named with SIGNED_BOT:.

Verified.
Personal Agent Protocol

Verified for vetted agents. Sets delegated to true.

Not documented.
Evidence reporting

evidence and spoofed on each decision, in your code.

High, medium, or low identification strength per agent, in dashboards. A request header for it isn't documented.

Bot catalog

Open source well-known-bots.

Not documented as public.
Policy

Allow or deny by bot ID or category in code, plus your own logic on the identity. Remote rules in the Console, CLI, or MCP.

Intent-Based, Allow, Block, Rate Limit, Time-Box, or Monetize, per agent, endpoint, or custom rule, in the dashboard.

Timeout behaviorHTTP checks can fail open, and you can configure that behavior.

Node.js module: allows the request after a 150 ms default timeout.

Pricing

Published: $25 or $299 per month, or custom, plus $5 per million web requests.

Plans linked to a number of requests.
Beyond bots

Rate limiting, Shield WAF, email validation, and policy on agent tool calls and coding-agent hooks.

Fraud protection products, including account protection and mobile SDKs.

Why does the decision's location matter for AI agents?

The decision's location matters because a verified agent identity doesn't settle what the agent may do. A signature proves which operator sent the request, and a Poppy session proves that a vetted agent acts for one of its users. Neither proves that the user approved this purchase, and neither knows the order total.

In DataDome, a security team sets a policy for each agent and endpoint, and your server acts on the module's result. In Arcjet, the identity arrives in the handler that already holds the session, the cart, and the plan. One function can allow a delegated agent to read the signed-in user's cart, require SIGNATURE evidence on checkout, and send an unverified "agent" to the API with a rate limit. For more information, see agentic commerce bot protection.

Neither product can identify an agent that drives a person's own browser, such as Claude in Chrome, because those requests look like the person. For more information, see how to detect AI browser agents.

When to choose which

Choose DataDome for a security team's dashboard across many integration points, Arcjet for bot decisions in code with application context, and both when you want each. The following sections list the signals for each choice.

Choose Arcjet

Choose Arcjet if any of the following apply:

  • You want bot decisions as code, reviewed and tested with the route they protect.
  • You want the evidence level and delegated flag in your handler, not only in a dashboard.
  • You want Poppy sessions verified alongside Web Bot Auth, in one identity.
  • You want an open bot catalog that you can read and contribute to.
  • You want published usage pricing.
  • You want rate limits, Shield WAF, and agent tool-call policy in the same SDK.

Choose DataDome

Choose DataDome if any of the following apply:

  • A fraud or security team owns bot policy and wants to change it without code.
  • You need one product across CDNs, web servers, and iOS and Android apps.
  • You want integration at the edge in front of your origin.
  • You want monetization policies for AI agents through DataDome's partners.

Use both

DataDome at the edge and Arcjet in the handler can run together. DataDome filters traffic before your origin, and Arcjet applies per-route rules that use the identity with the user and the plan.

DataDome alternatives

DataDome alternatives include edge bot managers, fraud platforms, and in-code detection. The following list covers the main options as of November 2026:

  • Arcjet: bot identity with evidence, Web Bot Auth, and Poppy verification in your request handler on any host.
  • Cloudflare Bot Management: edge bot scoring, verified bots, and AI crawler and agent policies for traffic through Cloudflare. For more information, see Cloudflare vs Arcjet.
  • HUMAN: AI agent trust levels and per-agent permissions with enforcers on a CDN or origin.
  • Akamai Bot Manager: edge bot detection from a CDN provider.
  • Vercel BotID: a browser challenge with verified-bot fields for projects on Vercel. For more information, see Vercel BotID vs Arcjet.

For a side-by-side view of all of them, see AI agent detection tools compared.

Frequently asked questions

What is the difference between Arcjet and DataDome?

DataDome integrates at your CDN, web server, or backend, with the edge recommended, and a security team sets bot and AI agent policy in its dashboard. Arcjet runs in your request handler and returns a bot identity your code reads: the kind of bot, its ID, the evidence that established it, and whether a verified Personal Agent Protocol session shows it acts for a user.

Is Arcjet an alternative to DataDome?

Yes, for teams that want bot decisions in code. Arcjet covers bot detection, AI agent identity, rate limiting, and Shield WAF with SDKs for JavaScript, Python, and Go, on any host, with published usage pricing. DataDome fits teams that want one dashboard across CDNs, web servers, and mobile apps. Arcjet doesn't absorb network-layer DDoS attacks, so pair it with a DDoS provider.

Does DataDome verify Web Bot Auth?

Yes. DataDome's bot authentication docs state that it can cryptographically verify HTTP signatures using the Web Bot Auth protocol, and its AI agent identification docs rate Web Bot Auth as a high-strength method. Arcjet also verifies Web Bot Auth, for the signers in its open well-known-bots catalog and any directory you name.

Does DataDome support the Personal Agent Protocol?

DataDome's documentation doesn't mention the Personal Agent Protocol (Poppy), as of November 2026. Arcjet verifies Poppy sessions for agents vetted in its well-known-bots catalog and reports them with SIGNATURE evidence, the agent's domain as operator, and delegated set to true.

Can you use DataDome and Arcjet together?

Yes. DataDome can filter traffic at the edge before it reaches your origin, and Arcjet can apply per-route rules in the handler, using the bot identity with the signed-in user, the cart, and the plan.

How is DataDome priced compared with Arcjet?

DataDome's documentation describes subscription plans linked to a number of requests, and log enrichment requires a Corporate or Enterprise plan. Arcjet publishes its prices: Individual ($25 per month), Startup ($299 per month), and custom Enterprise plans, plus $5 per million web requests.

AI runtime security in your code

Protect your AI agent workflows with Arcjet

Arcjet runs inside your application, where it can use runtime context to enforce agent actions and budgets, detect prompt injection, and protect sensitive information before a workflow acts.