Platform

Gaming

Resources

Enterprise

Company

Platform

Gaming

Resources

Enterprise

Company

Migration

/

AWS GameLift

AWS GameLift

Edgegap

Published

Published

Published

Last reviewed for accuracy

Last reviewed for accuracy

Last reviewed for accuracy

Migrate from AWS GameLift with zero impact on your players

Migrate from AWS GameLift with zero impact on your players

Remove the Server SDK, rebuild and containerize, and deploy on demand across 615+ locations. Your game logic doesn't change.

With GameLift you pre-provision fleets on one provider's regions and pay per core whether they're full or not. Edgegap's hybrid orchestration runs your steady load on egress-free bare metal Private Fleets and bursts to Edge Cloud on demand, across 17+ providers, so you get the cost of reserved capacity with the reach of on-demand, and no single provider outage can take your whole footprint with it.

Talk to an Engineer

No credit card required. Free tier includes the resources to deploy a real game server.

No credit card required. Free tier includes the resources to deploy a real game server.

What changes (and doesn't change)

Changes to expect, at a glance

Changes to expect, at a glance

A GameLift migration is a hosting change, not a rewrite. Here's what to expect from your migration.

A full feature-by-feature breakdown is in Step 3: Identify & Replace features below.

What Remains the Same

Your netcode and transport

Edgegap never touches your transport, serialization or tick loop. The container runs the same server binary you build today.

Containerize

Removing the SDK means one rebuild, which any game already in production is set up to do. Edgegap's automated build tools can fast track it, for Unity, Unreal and Godot.

Your client connect flow

Your client still receives an address and a port and connects to it. Only where that address comes from changes.

Engine, versions & CI

Same engine, same version, same build pipeline. Nothing in your CI needs a new step beyond pushing the image

What Changes

The Server SDK comes out

Five GameLift calls get removed; InitSDK, ProcessReady, ActivateGameSession, ProcessEnding and Destroy.

One allocation call replaces the fleet

Where you asked a fleet for a session, you POST once to the deploy endpoint and get back an FQDN and port in the same response.

Regions stop being a purchase decision

You no longer buy capacity region by region. Every deployment is placed at the best of 615+ locations at request time for Edge Cloud.

Capacity stops being your job

No warm capacity to size, no scaling policies to tune, no idle instances between peaks. Edge Cloud has nothing to pre-provision at all, and Private Fleets get rightsized with us as your traffic curve moves, so capacity stops being a thing you manage.

Expected Timeline

Five simple steps

Most studios reach end-to-end validation in under two weeks, including parallel running while traffic rolls out.

Most studios reach end-to-end validation in under two weeks, including parallel running while traffic rolls out.

Total: under 1 week

Total: under 1 week

Step 1

Remove the server SDK

Remove the server SDK; delete GameLift init and lifecycle calls.

~1 hour

Step 2

Containerize

Removing the SDK means one rebuild. Automated build tools are available for Unity, Unreal and Godot, or build the image yourself.

~30 min. to 2hrs.

Step 3

Push to a Registry

Edgegap's registry is available for free, or keep ECR, DockerHub, GitLab, etc.

~30 min.

Step 4

Swap the Allocation Call

One HTTP request replaces fleet placement.

~1 day

Step 5

Deploy Globally

Deploy globally on Edge Cloud, and optionally cut cost further with egress-free hosting on Private Fleets.

Instant Rollout

Detailed migration steps

1

The migration steps, in detail, with the code

If you are looking to migrate your multiplayer game from AWS GameLift to Edgegap's game server hosting & orchestration, the following simple steps will get you running in no time.

Make sure to refer to our "Switching from GameLift to Edgegap" documentation for the latest updates to this process.

Before getting started, we expect that:

  • You currently use AWS GameLift

  • You currently have a working game server build on GameLift

Step 1:

Remove the AWS GameLift SDK

Edgegap does not require a server SDK to integrate.

The first step to switch to Edgegap is to remove the code that initializes AWS GameLift within your game server. This is to reduce overhead and prevent errors. The code you are looking to remove probably looks like the following.

Unity (C#)
using UnityEngine;
using Aws.GameLift.Server;
using System.Collections.Generic;
 
public class GameLiftServerExampleBehavior : MonoBehaviour
{
    //This is an example of a simple integration with GameLift server SDK that will make game server processes go active on GameLift!
    public void Start()
    {
        //Identify port number (hard coded here for simplicity) the game server is listening on for player connections
        var listeningPort = 7777;
 
        //InitSDK will establish a local connection with GameLift's agent to enable further communication.
        var initSDKOutcome = GameLiftServerAPI.InitSDK();
        if (initSDKOutcome.Success)
        {
            ProcessParameters processParameters = new ProcessParameters(
                (gameSession) => {
                    //When a game session is created, GameLift sends an activation request to the game server and passes along the game session object containing game properties and other settings.
                    //Here is where a game server should take action based on the game session object.
                    //Once the game server is ready to receive incoming player connections, it should invoke GameLiftServerAPI.ActivateGameSession()
                    GameLiftServerAPI.ActivateGameSession();
                },
                (updateGameSession) => {
                    //When a game session is updated (e.g. by FlexMatch backfill), GameLiftsends a request to the game
                    //server containing the updated game session object.  The game server can then examine the provided
                    //matchmakerData and handle new incoming players appropriately.
                    //updateReason is the reason this update is being supplied.
                },
                () => {
                    //OnProcessTerminate callback. GameLift will invoke this callback before shutting down an instance hosting this game server.
                    //It gives this game server a chance to save its state, communicate with services, etc., before being shut down.
                    //In this case, we simply tell GameLift we are indeed going to shutdown.
                    GameLiftServerAPI.ProcessEnding();
                },
                () => {
                    //This is the HealthCheck callback.
                    //GameLift will invoke this callback every 60 seconds or so.
                    //Here, a game server might want to check the health of dependencies and such.
                    //Simply return true if healthy, false otherwise.
                    //The game server has 60 seconds to respond with its health status. GameLift will default to 'false' if the game server doesn't respond in time.
                    //In this case, we're always healthy!
                    return true;
                },
                listeningPort, //This game server tells GameLift that it will listen on port 7777 for incoming player connections.
                new LogParameters(new List<string>()
                {
                    //Here, the game server tells GameLift what set of files to upload when the game session ends.
                    //GameLift will upload everything specified here for the developers to fetch later.
                    "/local/game/logs/myserver.log"
                }));
 
            //Calling ProcessReady tells GameLift this game server is ready to receive incoming game sessions!
            var processReadyOutcome = GameLiftServerAPI.ProcessReady(processParameters);
            if (processReadyOutcome.Success)
            {
                print("ProcessReady success.");
            }
            else
            {
                print("ProcessReady failure : " + processReadyOutcome.Error.ToString());
            }
        }
        else
        {
            print("InitSDK failure : " + initSDKOutcome.Error.ToString());
        }
    }
 
    void OnApplicationQuit()
    {
        //Make sure to call GameLiftServerAPI.Destroy() when the application quits. This resets the local connection with GameLift's agent.
        GameLiftServerAPI.Destroy();
    }
}

2

Step2:

Containerize your game server

With the SDK removed, rebuild your game server and package it as a container image. A game already in production has the pipeline for the rebuild; the packaging is the only new part, and the guides below automate it for each engine.

Once containerized, simply push to a container registry Edgegap can pull from. Edgegap's own registry is free and needs no configuration; ECR, Docker Hub, GitLab and any OCI-compatible registry work unchanged.

3

Step 3:

Identify & Replace features

Swap the allocation call (CreateGameSession → DescribeGameSessions → GetConnectionInfo). That's it!

Optionally, review and identify services and features you currently use in AWS GameLift that may need to be replaced.

AWS Gamelift - Feature

Edgegap - Feature

Server Build Plugin

Engine native, fully automated quickstart plugins: Unity / Unreal Engine / Godot.

Server SDK

No server SDK required! ServerConfig replaced by Apps and Versions, managed without rebuilding.

Build Configuration

Fully automated. Optionally expose additional ports or variables.

Game Client Authentication

Optionally, add fine-grained access controls: matchmaker API token, server browser client/server API tokens + federated identity, custom DIY session management (Edgegap API token).

FlexMatch Self-Hosted

No-code managed matchmaker, test for free. See also the optional lightweight SDK for Unreal and Unity.

SearchGameSessions API

Managed server browser, test for free. See also the optional lightweight SDK for Unreal and Unity.

Lobby

Group Up for Matchmaking or Reserve Seat in Server Browser.

Fleets - Horizontal Scaling

Private Fleets with Persistence and Cloud Overflow. Optionally add later on, or default to Match-Bound cloud.

AB testing, Aliases, Queues

Configurable Profiles (Queues), Instance Metadata and Rolling Updates.

Server Analytics

Deployment Analytics and Matchmaking Analytics.

4

Step 4:

(Optional) Setting up Fleets

Edgegap also offers an optional feature similar to AWS GameLift's fleets, if that is something from before you switched to Edgegap that you would like to keep using. Discover egress-free hosting with our Private Fleets, our bare metal, persistent servers.

Pricing

Price Comparison

Discover how Edgegap's orchestration is cost effective versus your current AWS Gamelift bill.

Cost driver 1: cloud-only vs hybrid orchestration.

AWS orchestrates cloud instances only. Edgegap orchestrates cloud and bare metal, and bare metal means egress-free hosting.


AWS Gamelift

Edgegap

Orchestration

Cloud orchestration only

Hybrid orchestration (cloud + bare metal)

Locations

  • Cloud: 35 GameLift-available locations (26 regions + 9 Local Zones), purchased individually, per location.

  • Edge Cloud: deploy to all 615+ locations (on-demand) on Edge Cloud

  • Private Fleet: 58 locations available (purchased individually, per location)

Egress Free Hosting

Only available on Gen6 (more expensive)

yes, via Private Fleets

Cost Driver 2: AWS "traditional" fleet-based orchestration vs Edgegap's modern "just-in-time, on-demand, and usage-based" container orchestration.


AWS Gamelift

Edgegap

Model

Fleet-based billing

On-demand, per container deployment & billing

Co-tenancy

No

  • Yes. Multi-tenant on Edge Cloud, with CPU and memory reserved per deployment.

  • No noisy-neighbor

  • Lower overall vCPU rate than reserved capacity on equivalent cloud.

Cost driver 3: paying for capacity vs paying for sessions.


AWS Gamelift

Edgegap

Idle capacity

Billed.

Warm fleets run and bill between peaks.

Not billed.

Zero active deployments costs $0 on Edge Cloud.

Billing granularity

Per instance-hour, per region, purchased in advance

Per vCPU-minute, fractionable from 0.25 vCPU (smaller with volume), no commitment.

Pricing

Optimize costs and performance with hybrid orchestration

Optimize costs and performance with hybrid orchestration

Optimize costs and performance with hybrid orchestration

Edge Cloud

On-Demand

$0.00115

/ vCPU

/ MINUTE

Billed by the Minute, Only on Usage

Pay only when players play. 0 active servers = $0

Deploy game servers worldwide, per instance, to all 615+ edge cloud locations. Zero commitment.

Simply add a credit card.

615+ locations worldwide

$0.10/GB of monthly Egress

0.25 to 4 vCPU deployments

Compare features & pricing

Edge Cloud

On-Demand

$0.00115

/ vCPU

/ MINUTE

Billed by the Minute, Only on Usage

Pay only when players play. 0 active servers = $0

Deploy game servers worldwide, per instance, to all 615+ edge cloud locations. Zero commitment.

Simply add a credit card.

615+ locations worldwide

$0.10/GB of monthly Egress

0.25 to 4 vCPU deployments

Compare features & pricing

Private Fleet (Bare Metal)

Reserved Capacity

$0.0256

/ vCPU

/ HOUR

Billed Monthly, Always On

$280 per host/month (12 mo rate), or $350 monthly

Optimize the cost effectiveness of your game server with egress-free hosting of Private Fleets.

Persistent, available 24/7, exclusive to your studio.

58 locations worldwide

Egress Included

16 vCPU/host total (15 usable)

Compare features & pricing

Private Fleet (Bare Metal)

Reserved Capacity

$0.0256

/ vCPU

/ HOUR

Billed Monthly, Always On

$280 per host/month (12 mo rate), or $350 monthly

Optimize the cost effectiveness of your game server with egress-free hosting of Private Fleets.

Persistent, available 24/7, exclusive to your studio.

58 locations worldwide

Egress Included

16 vCPU/host total (15 usable)

Compare features & pricing

Your bill varies per rate and usage; which can vary per game type, player count, & more.

Your bill varies per rate and usage; which can vary per game type, player count, & more.

Model your usage with our estimate calculator ->

Model your usage with our estimate calculator ->

Edgegap's features & price: detailed comparison between Edge Cloud & Private Fleets

Edgegap's features & price: detailed comparison between Edge Cloud & Private Fleets

COMPARE

Free Trial

Edge Cloud

Private Fleets

BILLING & PRICE

Rate per vCPU

Free

$0.00115/vCPU/minute

$0.0256/vCPU/hour

Pricing Model

Free

On-demand (usage-based)

Reserved capacity

Billing

No credit card required

Yes, via Stripe

Yes, via Stripe

Commitment

None

None

1, 6 or 12 months (incl. discount)

Egress

N/A

$0.10/GB month

Incl. 6 TB per host, per month

FEATURES

Analytics & Monitoring

Billing Alarms

Click to Expand

28 more rows

Unlocks

Capabilities are unlocked when a credit card is added to the account.

China deployments

DDoS protection

Fully-managed clusters

Always-on deployments

COMPARE

Free Trial

Edge Cloud

Private Fleets

BILLING & PRICE

Rate per vCPU

Free

$0.00115 / vCPU / minute

$0.0282 / vCPU / hour

Pricing Model

Free

On-demand (usage-based)

Reserved capacity

Billing

No credit card required

Yes, via Stripe

Yes, via Stripe

Commitment

None

None

1, 6 or 12 months (incl. discount)

Egress

N/A

$0.10/GB month

Incl. 6 TB per host, per month

FEATURES

Analytics & Monitoring

Billing Alarms

Click to Expand

28 more rows

Unlocks

Capabilities are unlocked when a credit card is added to the account.

China deployments

DDoS protection

Fully-managed clusters

Always-on deployments

COMPARE

Free Trial

Edge Cloud

Private Fleets

BILLING & PRICE

Rate per vCPU

Free

$0.00115/vCPU/minute

$0.0256/vCPU/hour

Pricing Model

Free

On-demand (usage-based)

Reserved capacity

Billing

No credit card required

Yes, via Stripe

Yes, via Stripe

Commitment

None

None

1, 6 or 12 months (incl. discount)

Egress

N/A

$0.10/GB month

Incl. 6 TB per host, per month

FEATURES

Analytics & Monitoring

Billing Alarms

Click to Expand

28 more rows

Unlocks

Capabilities are unlocked when a credit card is added to the account.

China deployments

DDoS protection

Fully-managed clusters

Always-on deployments

Features Comparison

Features Comparison


Edgegap

AWS GameLift

Orchestration

Just-in-time & on-demand, instance-based

Traditional, fleet-based pre-provisioned capacity. Scale-to-zero added January 2026, with a cold-start delay when scaling back from zero.

Infrastructure

17+ vetted providers (cloud and bare metal)

AWS only. Single-provider; no cross-provider rerouting.

Regions

615+ Edge Cloud locations worldwide, 58 Fleet locations globally

35 GameLift-available locations as of July 2026 (26 regions + 9 Local Zones), each selected and billed separately

Pricing

$0.00115/min per vCPU (fractionable), $280/month Performance host (16 vCPU, 32 GB, 6 TB egress) / $320/month Overdrive (3.7-5.1 GHz, 64 GB) on a 12-month commitment

Per-region instance pricing, purchased location by location. Network bandwidth is free on generation-6+ instances as of June 2026 (excludes China); other add-ons may apply: Elastic IP, ELB, FlexMatch, CloudWatch, support plans.

Scalability

14M CCU with real world testing, 135M sessions LTD, automatic with no pre-provisioning

100M CCU tested via 83,333 pre-provisioned VMs; real-world peak ~14M CCU per AWS, requiring manual fleet pre-configuration

Integration

Unity/Unreal plugins + netcode & backends

Unity/Unreal plugins only; no netcode samples; Unreal still requires building from Source

Support

24/7 for clients, free. Available on Discord, Slack, email

No developer community or support channel highlighted, even as a paid option

Latency

58% reduction proven

States latency is "critical" but publishes no latency performance data for GameLift

Track Records

8+ years, 1,600+ studios, live AAA

Long-running AWS service; current case studies include Apex Legends and others (roster refreshed periodically, as of August 2026)

Every GameLift figure is cited from AWS documentation or announcements on the date shown. Full source list.

Clients

Learn more from clients who migrated to Edgegap

15 case studies are available from studios that moved to Edgegap from another platform.

Key reasons why they chose Edgegap

Key reasons why they chose Edgegap

Discover the impact of Edgegap's automated game server orchestration for a flawless multiplayer experience.

+58% Average Latency Reduction

+58% Average Latency Reduction

14M CCU Rapid-Scaling

14M CCU Rapid-Scaling

+615 Locations Worldwide

+615 Locations Worldwide

~2s. Median Game Server Boot Time

~2s. Median Game Server Boot Time

Multi-Version Support & Rolling Updates

Multi-Version Support & Rolling Updates

91% Fill Rate Matchmaking

91% Fill Rate Matchmaking

Cross-Platform Compatibility

Cross-Platform Compatibility

Instant DDoS Protection

Instant DDoS Protection

Multi-Cloud Across 17+ Providers for 99.99% Availability

Multi-Cloud Across 17+ Providers for 99.99% Availability

Methodology

How we measured

Each metric states what is measured, over what window, and what is excluded. Benchmark results are dated to the test that produced them and are not restated as current figures.

Orchestration Platform Scale

Where servers can run, and how many we have run.

Inventory

Global Edge Cloud locations

615+locations

Distinct deployment targets where a game server container can be scheduled on demand.

Platform inventory
How we measure this
Live

Private Fleet locations

58locations

Locations where a studio can reserve dedicated bare-metal or VM hosts for its exclusive use.

Published live
How we measure this
Cumulative

Game server deployments to date

135million deployments

Total game server deployments orchestrated since February 2019, including deployments that failed to start, the same basis our competitors use.

Since February 2019
How we measure this

Performance Under Load

How fast our orchestration delivers a server, and what it sustains under test.

Rolling 30d

Server deployment time

2seconds · median

From cold start: provisioned from scratch, not allocated onto an already-running server. Measured to container ready, before the game engine starts.

Rolling 30-day window
How we measure this
Benchmark

Sustained deployment rate for each game

40per second

Started deployments per second, averaged over a continuous 60-minute benchmark. Available to each game from confirmed studios with live traffic.

November 2023 benchmark
How we measure this
Benchmark

Concurrent capacity for each game

14million concurrent

Concurrent sessions reached on real game server containers, at 100 players per server.

November 2023 benchmark
How we measure this

Reliability

What we commit to contractually for our orchestration, and what the monitors record.

Committed

Availability

99.99%SLA commitment

Committed under our SLA. Trailing 90-day measured uptime: 99.999%.

Trailing 90 days
How we measure this
Inventory

Provider redundancy

17+providers

Cloud and bare-metal providers under contract and carrying production traffic. Failover is automatic.

Platform inventory
How we measure this

Live component status is published at status.edgegap.com.

How We Measure

One entry per metric: what is measured, how, and over what window.

Global Edge Cloud locations

A location is a distinct deployment target: one datacenter site at one of our infrastructure providers, in which a game server container can be scheduled on demand. Sites are counted once each regardless of how many providers operate in the same metro, and count only while accepting deployments.

Back to the metric ↑
Private Fleet locations

The subset of the 615+ where a studio can reserve dedicated bare-metal or VM hosts for its exclusive use rather than deploying into shared on-demand capacity. Published live on the Fleet Manager page. This is also the number that carries the included-egress allowance.

Back to the metric ↑
Game server deployments to date

Counted as total deployment records since February 2019, including deployments that failed to start, the same basis our competitors use.

Back to the metric ↑
Server deployment time

From cold start: provisioned from scratch, not allocated onto an already-running server. Measured to container ready, before the game engine starts.

Back to the metric ↑
Sustained deployment rate for each game

The orchestrator started real game server containers continuously for 60 minutes, averaging 40 started deployments per second over the full hour, counted at deployment start rather than at ready state. Measured across 550+ locations and 17 providers in November 2023. The benchmark ran for a single game. Capacity is allocated per game, so every game on the platform can scale to this level independently and at the same time.

Back to the metric ↑
Concurrent capacity for each game

In November 2023, the platform sustained 40 game server deployments per second for 60 minutes across 550+ locations and 17 providers, reaching 14 million concurrent sessions at 100 players per server. Game servers were real containers; player clients were simulated. The benchmark ran for a single game. Capacity is allocated per game, so every game on the platform can scale to this level independently and at the same time.

Back to the metric ↑
Availability

Availability is committed at 99.99% under our SLA and measured continuously against five independently monitored components: Container Registry, Dashboard, Public API, Public Documentation and Website. Trailing 90-day measured uptime: 99.999%. The live figure is published on status.edgegap.com.

Back to the metric ↑
Provider redundancy

Distinct suppliers under contract and carrying production traffic, counted at corporate-entity level. Failover across providers is automatic; it is the premise of the multi-cloud architecture.

Back to the metric ↑

How We Keep This Current

What gets reviewed, on what cadence, and what carries its own date.

Platform figures on this page are reviewed and refreshed quarterly. Metrics drawn from live sources (availability and game server deployments) carry their own as-of dates and are updated as measured. Benchmark results are dated to the test that produced them and are not restated as current figures.

Last updated
Next reviewor earlier

Frequently Asked Questions

Edgegap is the hosting & orchestration experts. Our platform is tailor built to help you speed up your development process, which allows you focus on what you excel at - development.


Do I have to rewrite my netcode?

No. Edgegap does not touch your transport, netcode, or serialization — the container runs the same server binary you build today. The migration removes the GameLift Server SDK and changes one backend call: where you asked a fleet for a game session, you now POST to the Edgegap deploy endpoint and get back an FQDN and port. Your client connects to that address exactly as it connected to a GameLift-assigned one.

How long does a GameLift migration take?

Most studios reach end-to-end validation in under two weeks. Removing the SDK takes about an hour, containerizing takes 5 to 20 minutes with the Unity quickstart plugin or Unreal's build-from-container path, pushing to a registry about 30 minutes, and swapping the allocation call about a day. The remaining time is parallel running — keeping GameLift live while you shift traffic region by region.

Do I need an Edgegap server SDK?

There is no Edgegap server-side SDK, so there is nothing to integrate. The container starts, reads its injected port from the environment, and listens. What GameLift expressed as ServerConfig and lifecycle callbacks is configuration in Edgegap: Apps and Versions, set once in the dashboard or over the API.

What happens to my FlexMatch rule sets?

FlexMatch rule sets do not transfer, but you have two paths. Either rebuild the rules in Edgegap's no-code matchmaker, which is free to test and ships lightweight client SDKs, or keep your existing matchmaker and call Edgegap only for allocation once a match is formed. The second path is the shorter one if your rule sets are complex or heavily tuned.

Can I keep using my AWS credits?

Credits apply to AWS infrastructure, and AWS is one of the 17+ providers Edgegap deploys onto. Credits are drawn against AWS billing rather than Edgegap invoices, so the arrangement depends on your agreement with AWS. Worth noting that credits cover the bill, not the architecture — the orchestration model is what determines the bill once they expire, which is the case for migrating while you still have runway.

Does Edgegap run on AWS infrastructure?

Yes. AWS is an Edgegap partner and one of 17+ cloud and bare-metal providers under the orchestrator. The difference is that you do not select or pay for AWS regions individually — the orchestrator places each deployment at the best available location of 615+, and fails over across providers automatically.

Get your Game Online Easily & in Minutes

Start Integrating Now!

Get your Game Online Easily
& in Minutes

Get Started
Get Started
Get Started
Get Started
Get Started
Get Started