PlayFab's New Free Tier: 1,000 Players Unless You Ship on Xbox

On March 11, 2026, Microsoft introduced Foundation Mode for PlayFab at GDC. The dev-community headline that followed - "PlayFab cut its free tier from 100K to 1K" - is the version we originally ran here too, and it isn't quite right. PlayFab's pricing page has long carried a "$0/month before your title passes 100K players" marketing line, but that was never the Development Mode account limit. The accurate version is more specific and, for non-Xbox indies, not much better: the broad $0 path now requires Xbox, and the evaluation path caps at 1,000 lifetime accounts.

The PR framing is that Foundation Mode is "free" - and it is, if you check three boxes. This post explains the boxes, what happens if you can't check them, and what indie studios who don't ship on Xbox are actually doing.

What changed, in one paragraph

PlayFab's evaluation path used to be positioned as a no-friction sandbox with a "$0 before 100K players" headline on the pricing page. As of the March 2026 restructure, Microsoft's plan table is explicit: Development Mode supports up to 1,000 lifetime player account creations per title, across up to 10 development titles, with no payment instrument required. Above that, you either qualify for Foundation Mode, move to pay-as-you-go or a paid tier, or you leave.

Foundation Mode: free, with strings

Foundation Mode is the new free path, but it's gated on three eligibility checks:

1. You register a studio in PlayFab Game Manager.
2. You ship or plan to ship your game on Xbox.
3. You link your PlayFab account to your Partner Center studio and product.

Hit all three and Foundation Mode gives you PlayFab's core cross-platform services at no additional cost: identity, progression, community, multiplayer, live service management, economy, and game data stream. No payment instrument required.

Two catches. The first is step 2: if your game is PC-only, mobile-first, Switch-exclusive, or otherwise not Xbox-bound, Foundation Mode is not for you, and you're back in Development Mode's 1,000-account corral. The second is that Foundation Mode is not all of PlayFab: Microsoft's mode overview lists PlayFab Multiplayer Servers, UGC, classic CloudScript, Segmentation, Experimentation, CDN, and several other services as blocked in Foundation Mode, and Azure Functions compute bills separately through Azure.

What 1,000 lifetime accounts actually buys you

1,000 sounds like a number until you do the math on a real playtest.

Phase	Typical account creation	When the wall hits
Internal alpha	~20-50 accounts (team + ops)	Comfortable
Friends-and-family playtest	~50-200 fresh accounts	Mostly comfortable
Discord closed beta (200 testers)	~1,000-3,000 (re-tests on fresh accounts)	Hits the wall mid-beta
Steam playtest / Next Fest demo	~10,000+ in week one	Wall hit immediately

The Development Mode cap is fine for an internal sandbox. It's not fine for any external playtesting larger than the team's group chat.

What indie studios are switching to

Studios who looked at the Foundation Mode requirements and concluded "this isn't for us" have a small handful of real options in mid-2026.

Crux

The free Dev tier covers 10,000 monthly active users with every feature included - auth, persistent player data, leaderboards, server registry, live config, and matchmaking. No Xbox requirement, no Partner Center linkage, no payment instrument, and it never auto-pauses. Paid plans above it are flat-rate ($10/month Launch, $79/month Scale) instead of usage-based meter math.

Best fit for: indie / mid-size studios who want a unified backend with predictable pricing and no platform lock-in. See pricing.

Nakama (self-hosted)

Open source, Apache-licensed, you run the ops. Infinite "free" in the sense that there's no per-account cap, but you're now responsible for PostgreSQL, your Linux fleet, your backups, and your on-call rotation. Heroic Labs offers a managed Nakama Cloud if you want to outsource the ops.

Best fit for: studios with a platform engineer on staff who would rather own the infrastructure than rent it.

Firebase

Google's BaaS. Free tier is generous for auth and Firestore reads. The gap: Firebase has no dedicated game server orchestration, no built-in leaderboards (you build them on top), no live ops dashboard for game-specific events. It's a good auth+database foundation but not a complete backend for multiplayer games.

Best fit for: turn-based mobile games and casual multiplayer where dedicated servers aren't needed.

Beamable

Managed game backend with a free tier and per-MAU paid pricing. The free tier is smaller than Crux's but the platform is mature and has been live for several years. Strong on live ops tooling specifically.

Best fit for: studios who want a managed alternative to PlayFab without re-onboarding to Microsoft's platform.

Migration: what touches and what doesn't

If you're migrating off PlayFab, the work is real but bounded. Here's what your engineering team actually touches:

• Client SDK calls - every PlayFab.AuthenticationApi.LoginWith*, PlayFab.LeaderboardsApi.*, PlayFab.DataApi.* call becomes a different SDK call. Wrap PlayFab in your own thin client BEFORE migrating to make this easier.
• CloudScript handlers - server-authoritative game logic running in PlayFab needs a new home (your own backend service, or the equivalent feature in your new BaaS). Note that classic CloudScript is one of the services Foundation Mode blocks anyway.
• Scheduled tasks - PlayFab's scheduler runs cron-style jobs. Whatever replaces it (cloud cron, k8s CronJob, your new BaaS scheduler) needs to be wired up before you cut over.
• Player ID continuity - the hardest piece. If existing players have PlayFab IDs and saved data, you need a one-time export and mapping into the new backend. Without this, your live game's players lose their progression on migration day.
• Live ops dashboard - your community manager / data team currently watches PlayFab's dashboard. Whatever they watch next needs the same event feed wired up before cut-over.

Two to four engineering weeks is the standard estimate for a mature title. Most of the time is in the player ID migration and validation, not the API call rewrites.

What we're recommending in August 2026

If you were evaluating on PlayFab's old terms and the March restructure caught you out, here's the decision tree:

• Shipping on Xbox in the next 12 months: Foundation Mode is genuinely free for the included services and the path of least resistance. Check the blocked-services list first (no Multiplayer Servers, no UGC, no classic CloudScript), then sign up, link Partner Center, move on.
• Not shipping on Xbox, indie / mid-size: Crux. The free Dev tier covers a real playtest, no platform gating, and the paid plans are flat-rate instead of usage-based meter math. See pricing.
• Have a platform engineer and want to own the stack: Self-hosted Nakama. Free in dollars, expensive in ops time. Worth it if your team has the muscle.
• Casual / turn-based mobile only: Firebase. It's not a game backend, but for non-realtime mobile titles it's a fine foundation.

Related reading

• PlayFab Alternatives for Indie Games in 2026 - LootLocker, Nakama, BrainCloud, and Crux compared for non-Xbox indies.
• PlayFab vs Crux - the deeper feature-by-feature and pricing-model comparison.
• Unity Multiplay Shut Down: Where Studios Actually Went - if your PlayFab setup also used Multiplay for hosting, you need to plan both migrations at once.
• Game Server Backend Platform Buyer's Guide - broader buying framework.
• Cost calculator - model your MAU and API call curve against Crux's flat-rate pricing.