Bounty Board Arcade vs CrazyGames, Poki, and Newgrounds

Ask almost anyone where to put a finished browser game and you get the same ladder back: itch.io first, then Newgrounds, then CrazyGames, then Poki.

The order is roughly right. What it hides is the question that matters once the build is done: which of those gates will take your game, and what happens when one opens.

They are not the same kind of gate. One is a public vote. One is a quick QA pass followed by a data test on live traffic. One is a measured funnel with five levels before anyone mentions a contract. Here is what each asks as of August 2026, from the platforms' own documentation, plus where Bounty Board Arcade honestly sits. Money is a separate subject, covered in how to monetize an HTML5 game.

The direct answer

• Newgrounds: publish free, live immediately, and the community votes it through or deletes it.
• CrazyGames: easy to enter, hard to graduate from. A Basic Launch needs no customization, then the game earns a Full Launch on live metrics.
• Poki: slowest and strictest. Closed beta access and a measured playtime bar. Poki prefers web-exclusive deals that default to 5 years, but it also publishes a non-exclusive flat-license option.
• Bounty Board Arcade: a human review, most completing within 2 business days. Much smaller audience, but leaderboards and SDK services are free, eligible studios take an ad revenue share, and creator campaigns run from the same account.

Newgrounds: the crowd is the gate

Newgrounds publishes an HTML5 upload immediately and uses community judgment instead of a pre-publication staff review.

Then it enters judgment. Registered users vote from 0 to 5 while the score and the reviews stay hidden from everyone but the reviewer and the author. Fall below a threshold and the submission is deleted automatically, which the site calls getting blammed. The community wiki records the current thresholds as 1.0 out of 5 at 40 votes, 1.5 at 50, and 2.0 at 60. Clear all three and the game stays permanently. It is a real gate, just an unusual one: it punishes a weak first thirty seconds far harder than a shallow game, because most voters never reach the depth.

Passing buys more than its reputation suggests. Written reviews from people who play a lot of browser games are hard to get anywhere else, and Newgrounds.io, the site's free developer API, has long offered medals, scoreboards, account sign in, and event tracking. If a leaderboard is all you want, it is already sitting there for nothing. GDevelop's guide adds the other argument: Newgrounds "promotes new games organically, rather than only promoting those that pay to be promoted".

Aim here if you want fast, blunt feedback on a jam game, a prototype, or something too strange for a portal. It is a feedback surface with a loyal community, not a growth engine.

CrazyGames: get in quickly, then prove it in the data

CrazyGames is the friendliest front door of the three portals and the most honest about where the real test happens. The publishing process runs in two stages. A Basic Launch can go live "without needing to customize your game for CrazyGames": the SDK is optional and monetization is off. A Full Launch requires the SDK, and ads run through the platform. The technical requirements are generous by portal standards: initial download 50MB or less, 250MB total, 1500 files, and 20MB or less for mobile homepage eligibility.

The gate is what happens after you are live. The Basic Launch window ends once the game has been live at least 7 days and reached 500 plays, and closes automatically after 21 days if the plays do not arrive. CrazyGames publishes what a good result looks like: 10 or more minutes of average play time, 10% to 15% Day 1 retention, 80% or more of players converting, a load under 10 seconds, and a build below 20MB. The documentation frames those as "what success looks like" rather than hard pass marks.

There is no exclusivity. The FAQ states you can publish "even if it is already live on other platforms, including mobile app stores or PC storefronts, as long as you hold the necessary distribution rights." The tradeoff arrives at Full Launch instead, where the SDK goes in and monetization runs through them.

Aim here if your first minute is strong and you can act on retention data quickly. Getting listed is not the achievement. Surviving the 500 plays is.

Poki: five levels before anyone mentions a contract

Poki is the most selective of the four platforms here, and its process is unusually well documented. Access is a closed beta you apply to through a submission form, and Poki says outright that it "cannot respond to every submission but if your game is a good fit, we'll reach out to you."

Level 2 is playtesting: each test returns video recordings with button inputs and console output, and you must watch at least 10 before the next stage unlocks. Level 3 is the Player Fit Test, an automated run against 500 players that usually completes in about 5 hours. The bar is specific: an average overall playtime over 3 minutes, and at least 25% of the 500 gameplays over 3 minutes.

Level 4 is the Web Fit Test, about 3 to 5 days in front of around 10,000 players, scored 0 to 5 on click through rate, average time on page, and conversion to play. A 3 or higher means the game performs at or above the average game in its categories, and it goes automatically to the final review. Then the release process starts: 1 to 2 weeks of scheduling, a soft release of 2 to 3 weeks that cannot be skipped, and a global release typically 2 to 3 weeks after that.

Poki's preferred deal is the largest ask here. Its current deal types page describes web-exclusive contracts that default to 5 years and cover the open web, with Steam, mobile stores, and consoles left open but Discord and YouTube Playables counted as web. It also publishes a non-exclusive alternative: a one-time flat license with no revenue share or marketing investment. The requirements shape the build too: target an initial download under 8 MB, scale to a 16:9 canvas, and run no ads other than Poki's own, because "Poki's SDK comes out of the box with advertisements."

Credit where it is due on rejection. Poki's guidance for reading test results says it is "not the end of your game's journey on Poki. Neither is it a signal from Poki that your game is not good," and further tests stay free while you iterate.

A studio writeup from Kuyi Mobile in February 2026 describes the same funnel from the outside, with numbers matching the documentation: "125 out of 500 players (25%) must play the game for at least 3 minutes," then a larger pool, then a soft launch of at least two weeks. Their line about the contract moment is the useful part: "If your game does well, the Poki team will reach out." The gate is not a taste decision you can pitch past. It is a measured playtime bar.

Aim here if the game is built for a mass casual web audience and portal traffic is the business model. If it is a demo, a slice, or a marketing asset for a paid title, compare the non-exclusive license instead of assuming the preferred web-exclusive route fits.

Where Bounty Board Arcade sits

Smaller. That is the first honest thing to say. The Arcade's audience is a fraction of a portal's, so it is a second surface with a different job, not a substitute for a portal push.

The gate is a person, and it is fast. A studio account uploads a build zip with index.html at the root, uncompressed, up to 64 MB per file, 250 MB total, and 160 files, or links an HTTPS build you already host and keeps its own domain. Most reviews complete within 2 business days and you get an email either way, including a rejection. Updated builds are reviewed again while the live version keeps serving.

What passing gets you is the services, not the traffic. A curated Arcade page with play tracking, plus leaderboards, cloud saves, lifecycle events, and A/B experiments through the free SDK. Rewarded ad breaks are allowlisted during review for approved games hosted by Bounty Board, and multiplayer is reviewed and enabled per game. No SDK is required to submit in the first place.

It asks for nothing exclusive. No fee to publish and no requirement to pull another platform's build. On advertising, eligible Arcade studios keep 90% of finalized ad revenue attributed to their game, settled monthly via Stripe once the $50 minimum is reached. The category norm sits nearer half, with the published numbers in our monetization guide.

The part that sets Bounty Board apart here is the campaign bridge. The same studio account can hire creators to make content about the game, pointed at a page players can already play. Bounty Board charges a standard 10% platform fee on each side of completed work, and only on completed work: no subscription, no credits, no listing fee. Standard campaigns cost nothing up front because the studio is charged when it accepts work, while competitions and results-based campaigns fund their budget up front through Stripe, held by Bounty Board and refunded automatically if unused at close.

Two limits belong right here. View based bonuses and results based payouts are measured automatically only for YouTube and Twitch, so work on TikTok, Instagram, X, and websites is reviewed and approved manually. Bounty Board had 261 creator profiles as of August 27, 2026.

The order we would run

1. Newgrounds and itch.io on day one. Free, instant, no exclusivity, and feedback within days. Our roundup of web game platforms covers the rest of that tier.
2. CrazyGames next. Entry is cheap, and the Basic Launch window is the most useful free retention test a small team can get.
3. Bounty Board Arcade in parallel. Two business days, a leaderboard, an ad share, and a page a campaign can point at. It complements a portal push rather than replacing one.
4. Poki last, once your own numbers clear a 3 minute average and you have decided whether its preferred web-exclusive deal or its non-exclusive flat license fits the game.

If none of those has players yet, the gate is not your problem. Start with how to get your first 1000 browser game players, then come back with the retention data that gets you through a portal.

The Arcade takes builds whether or not a portal has said yes, and submitting one takes a zip and a review.

Sources worth reading

• CrazyGames: publishing requirements introduction
• CrazyGames: Basic Launch, the metrics that matter
• CrazyGames: technical requirements
• CrazyGames: FAQ and contact
• Poki: working with Poki
• Poki: Player Fit Test
• Poki: Web Fit Test
• Poki: release process
• Poki: requirements
• Poki: revenue deal types
• Poki: reading Web Fit Test results
• Newgrounds.io: the free indie game developers' API
• Wikigrounds: Under Judgment
• GDevelop: how to publish your game on Newgrounds
• Kuyi Mobile: my game production process, and how I got published on Poki