Asset 20 8 2
Does AI recommend your business? Run the free check →

Join 15,000 business owners, marketers and entrepreneurs. The Sunday newsletter you'll be annoyed only arrives once a week.

Article

10 Permit.io Alternatives for Fine-Grained Access Control

Permit.io alternatives differ less by brand than by where you want the permission check to run: inside your code as a library, in a hosted service you call over the network, or in a policy engine you deploy next to your app. Open-source engines cost nothing to license but you run them; managed services charge by user, request or active principal and run themselves.

Most teams shopping for a Permit.io alternative assume they're comparing products. They're not, or not at first. They're choosing where an access decision gets made, and in what kind of language the rules get written. Get them wrong and you'll be rewriting your permission model in eighteen months, whichever logo is on the invoice.

Where the check lives, and how the rules are written

An embedded library. The rules load into your application process and the check is a function call. There's no network hop, nothing extra to deploy and nothing to page anyone about at 3am. The catch is that every service in every language needs its own copy, and keeping the rules identical across them becomes your job. For a single monolith with role-based access, this is often all you need.

A hosted service. You send a question (can this user do this to that resource?) to someone else's API and get an answer back. You get a central place to change rules, an audit trail and no infrastructure to run. You also get a network call on the hot path, a bill that grows with usage and a dependency you can't fix yourself when it's slow. Check the vendor's regions against where your app runs before you fall in love with the console.

A policy engine you deploy. The middle road: an open-source engine runs as a sidecar or small service inside your own network. Decisions are fast because they're local, the rules live in one place, and you keep your data. In return you own uptime, upgrades and the job of syncing user and resource data into the engine. Some vendors here sell a managed control plane on top of the open-source engine, which is a sensible halfway house once you've got more than a handful of services.

The second fork is how you describe access. Policy-as-code tools have you write rules in a language: Rego for one well-known engine, Cedar for the AWS service. Rules read like logic: allow if the user's department matches the document's department and it's a weekday. That's excellent for attribute-heavy decisions and for anyone who needs to show an auditor a readable rule file.

Relationship-based tools follow the Zanzibar pattern Google described for its own systems. You store facts such as "Ana is an editor of folder 12" and "doc 88 sits in folder 12", and the engine walks the graph to answer "can Ana edit doc 88?" That's the natural fit for anything that looks like Google Drive: sharing, nested folders, teams inside organisations. The cost is that you now have a permissions database to keep in sync with your main one.

Open-source engines and libraries you run

OpenFGA

A CNCF open-source, relationship-based authorization engine with a readable modelling language for describing who can do what.

AuthZed

Teams wanting Zanzibar-style permissions use its open-source SpiceDB database, self-hosted anywhere or run for you in AuthZed's managed cloud.

Topaz

An open-source authorization service that blends policy ideas from OPA with Zanzibar-style relationships for apps and APIs.

Apache Casbin

Embedded library covering ACL, RBAC and ABAC models, a good match for a single application that wants access rules in-process.

Open Policy Agent

A general-purpose policy engine for managing rules across your stack, with every decision written in its Rego language.

Managed services and hybrids

Cerbos

For teams who like policies as code but want a hosted Hub to manage them, this engine handles apps, gateways and AI agents.

Auth0 FGA

Okta's hosted fine-grained authorization service, built on OpenFGA and aimed at multi-tenant B2B apps and APIs.

Amazon Verified Permissions

If you're already deep in AWS, this managed service evaluates Cedar policies for each authorization request your app sends.

Aserto

Small hosted service for resource-level access control across applications and APIs, managed from one central place.

Ory Permissions

Part of the wider Ory identity platform, this permissions service runs on Ory's network or on your own servers.

Where I'd start, by the shape of your product

Single app, simple roles, one language: an embedded open-source library. It's the least to run and the easiest to throw away later if you outgrow it.

A product built around sharing, folders and teams: a relationship-based engine, self-run if you have a platform team, hosted if you don't.

Work with me

Want AI doing the heavy lifting in your marketing?

I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.

Regulated data, auditors and attribute rules: policy-as-code, ideally with a managed control plane so rule changes are reviewed and logged.

Heavy AWS shop: the native managed service first, because fitting the rest of your stack tends to matter more than small feature gaps.

Sharing and compliance rules in the same product: a tool that blends relationships with policies, and a steeper first month while the team learns both ideas.

A three-rule bake-off you can run this afternoon

Don't trial these tools on toy examples. Take three rules your product already enforces, or wants to, and write each one in every candidate. Pick them so they pull in different directions:

Rule one, a plain role: admins of a workspace can delete any project in it. Every tool handles this. What you're measuring is how many lines it takes and whether a new hire could read it.

Rule two, a share: a user given a link to one document can view that document and nothing else in the folder. This is the case the relationship model was designed for, so watch how much extra work the policy-first tools need.

Rule three, a condition: a contractor loses edit rights the day their contract end date passes. This is attribute logic, the home ground of policy languages, so watch how the relationship-first tools express it.

For each tool, write the three rules, load a few thousand fake users and resources, and call the check from your own code. Note three numbers: lines of model or policy, time to first correct answer, and how long a check takes from your app's region. Then break something on purpose. Delete a user upstream and see how long until access disappears. That gap is where real incidents hide.

You'll usually find one tool wins rule two, another wins rule three, and the right choice is the one whose weak rule you can live with.

What drives the bill

There's no fair price range for this category, and I'd be wary of anyone who gives you one. Five of the ten tools publish a price, and no two of those five charge in the same unit. On the vendors' own pricing pages this month, one charges by monthly active principals, one by user each month, one per API request, one by the hour for a self-service deployment and one by the year. Free means different things too: three of the managed or hybrid services show a free tier on their pricing pages this month, each capped, and five tools describe themselves as open source, which is free to license but not to run.

Work out which unit hurts you. A consumer app with millions of occasional users can be cheap per request and ruinous per user. An internal tool with a few hundred staff making thousands of checks a minute is the reverse. Model your real traffic for a month in a spreadsheet before you take any pricing page at face value, and remember that self-running an open-source engine swaps a licence fee for engineering hours and on-call cover.

Do you make authorization tools or permission services? Tool makers can ask to be considered for a featured place on this list.

Mistakes that bite after launch

  • Syncing user and resource data late. If the engine learns about a removed user from a nightly job, that person keeps access until morning. Push changes as they happen, and measure the gap.
  • One check per item on a page load. A list of two hundred documents shouldn't mean two hundred network calls. Use the batch or list-filter call the tool offers, or ask it which items the user can see.
  • Folding login and permissions into one model. Who someone is and what they may touch change at different speeds and for different reasons. Keep them as separate systems with a clean hand-off.
  • Scattering checks through the code. If the rule is called from forty places in forty ways, no tool swap or audit will be easy. Put every check behind one function you own.
  • Never testing the denials. Teams test that admins can do things. Write tests that a contractor, a removed user and a user from another tenant can't.

Keep it portable: two names that changed hands

Two names you might have expected here have changed owner. Oso's site now says it's been acquired by Tiny and leads with an AI agent security platform, and Permify's site says it's been acquired by FusionAuth. They're left off this list for that reason. If either is already on your shortlist, ask about the roadmap and the contract terms before you sign anything long.

It's also worth keeping authentication and authorization separate in your head and your architecture. Logging people in is a different job, covered in my guide to the best authentication APIs. And if AI agents will act for your users, they need permissions too: the AI agent integration platforms you connect will call your APIs, and your access rules have to cover them.

The lesson isn't to avoid small vendors. It's to keep your permission model portable. Store relationships and policies in a format you could export, keep the check behind one function in your own code, and a change of owner becomes a weekend of work rather than a rewrite.

Want your tool on this list? Featured spots and reviews here are paid partnerships. Here's how to partner with me.

Published and maintained by the Lilach Bullock team, covering marketing, AI and business growth.
Your buyers are asking AI who to use. Does it say you?

See for free whether ChatGPT, Claude, Perplexity, Gemini and Google name you, and get the plan to become the answer.

Check my AI visibility →
Sundays only

Get the Sunday newsletter.

One email a week. AI experiments, marketing tactics, and the workflows Lilach is building right now in her own business.

Subscribe free

Let’s get your marketing running on AI.

Book a free 30-minute call

We figure out what you need, where AI fits in, and what working together would look like.

Book the call →

Or take the 30-second calculator

You’ll see the hours and the money quietly leaking out of your week, and the three workflows worth building first.

Take the calculator →

Or grab the free AI resource library

Prompt packs, templates, checklists, and swipe files. The exact tools I build for paying clients. Yours, free.

Get the library →
Keep reading

More from the blog.