Best rug checkers in 2026: we tested 4 on the same 122 tokens
Most “best rug checker” lists compare features. We pasted the same tokens into RugCheck, GoPlus, Honeypot.is and our own checker, and wrote down what each one said: 24 long-lived tokens that should come back clean, and 98 tokens launched in the last two days, the ones people actually check. We make one of these tools, so the rules we used to read every answer, ours included, and the raw data are on this page.
The short answer
- No single checker is enough. Of 98 brand-new tokens, 38 got a danger rating from at least one tool, and on 14 of those another tool said OK. Check every token in two tools and go with the worse rating.
- Solana: RugCheck plus a second opinion. RugCheck answered all 40 new Solana tokens; its warnings are mostly about thin liquidity and concentrated holders, and it rated Raydium’s RAY “danger” for the same reason. Unhosted looks at how a token was launched and traded instead (wallet farms, bundled launches, pulled liquidity); the two gave the same rating on 21 of 40.
- Ethereum, Base and BNB Chain: GoPlus for coverage, Honeypot.is for the sell test. GoPlus answered all 58 new EVM tokens and lists every owner power, but marks some legitimate powers red (LDO, VIRTUAL). Honeypot.is answered only 14: for the rest it found no pool it supports. It is the only tool here that also test-sells from real holders’ wallets, and that check called Trust Wallet’s TWT a honeypot.
- Unhosted (ours) answered all 122 tokens on all four chains and gave no false alarm on the 24 long-lived ones. Its weak spot: its test sell is stored, not live. On a new Ethereum token whose pool had already been drained it said caution where Honeypot.is said honeypot. We have since fixed that: an empty pool that a second source confirms, or a failed test sell, now reads red. The ratings on this page stay as they were on the test day.
How we tested
| Tools | Unhosted, RugCheck, GoPlus, Honeypot.is: every free checker with a public API we could call without a key |
| Chains | Solana, Ethereum, Base, BNB Chain; each tool only on the chains it covers |
| Long-lived tokens | 24: six per chain, years old and widely held (BONK, WIF, PEPE, UNI, DEGEN, CAKE…) |
| New tokens | 98: from GeckoTerminal’s busiest, trending and newest pools, pool under 48 hours old, 25+ buyers in a day |
| When | 21 October 2026, every tool within the same hour |
| Not tested | Token Sniffer: its site blocks automated checks and its API is paid |
Every tool answers in its own words, so we read each one the way its own site does. RugCheck prints GOOD, WARNING or DANGER from its score; GoPlus counts “risky” and “attention” items; Honeypot.is gives a risk level; Unhosted gives a headline rating. We copied the first two rules from the tools’ own page code on 21 October 2026 so a token reads here exactly as it reads there:
| Tool | Danger | Caution | OK |
|---|---|---|---|
| RugCheck | DANGER (score 5,000+) | WARNING (score 1,000–4,999) | GOOD (under 1,000) |
| GoPlus | any “risky” item: honeypot, 50%+ tax, closed source, owner can change balances, fake token… | “attention” items only: proxy, mintable, blacklist, 10–50% tax, hidden owner… | no items |
| Honeypot.is | honeypot, very high or high risk | medium risk | low or very low risk |
| Unhosted | red: Scam, High risk, Can’t sell, or a 30%+ sell tax | amber: Caution, Red flag, Disputed, a 10–30% sell tax, Watch insiders | green: No red flags; and grey answers from a finished check (Not rated on Solana, a fixed small transfer fee) |
“No answer” means the tool returned nothing usable for that token: not found, not analysed yet, or an error. We count it separately; it is not a pass.
We make one of these tools
Unhosted is ours. We picked the sample from a source none of the tools controls, called our checker the way anyone can (the public rug checker), and publish every answer, including the ones that make us look worse. If you think a rating is read wrong, the raw ratings and reasons are in the dataset below.
Test 1: do they cry wolf on 24 safe tokens?
These tokens have traded for years and are widely held. A “danger” here is a false alarm; a “caution” is usually a true note about the contract (a blacklist function, a mintable supply) that does not make it a scam.
| Tool | Chains | Tokens | Danger | Caution | OK | No answer |
|---|---|---|---|---|---|---|
| Unhosted | Solana, Ethereum, Base, BNB Chain | 24 | 0 | 0 | 24 | 0 |
| RugCheck | Solana | 6 | 1 | 0 | 5 | 0 |
| GoPlus | Ethereum, Base, BNB Chain | 18 | 2 | 9 | 7 | 0 |
| Honeypot.is | Ethereum, Base, BNB Chain | 18 | 1 | 0 | 17 | 0 |
| Tool | Token | Rating | What it said |
|---|---|---|---|
| GoPlus | LDO (Ethereum) | Danger | risky: owner can change balances; attention: trading can be paused; attention: anti-whale limit; attention: anti-whale limit can change |
| GoPlus | VIRTUAL (Base) | Danger | risky: owner can change balances; attention: mintable; attention: hidden owner |
| Honeypot.is | TWT (BNB Chain) | Danger | honeypot; high fail rate; tax buy 0% / sell 0% |
| RugCheck | RAY (Solana) | Danger | danger: Top 10 holders high ownership; warn: Single holder ownership; warn: Missing file metadata; warn: Mutable metadata |
| GoPlus | PEPE (Ethereum) | Caution | attention: trading can be paused; attention: anti-whale limit; attention: anti-whale limit can change; attention: blacklist |
| GoPlus | MOG (Ethereum) | Caution | attention: anti-whale limit; attention: anti-whale limit can change; attention: tax can change; attention: external call |
| GoPlus | DEGEN (Base) | Caution | attention: mintable |
| GoPlus | BRETT (Base) | Caution | attention: whitelist |
| GoPlus | AERO (Base) | Caution | attention: mintable |
| GoPlus | WELL (Base) | Caution | attention: proxy |
| GoPlus | Cake (BNB Chain) | Caution | attention: mintable |
| GoPlus | FLOKI (BNB Chain) | Caution | attention: external call |
| GoPlus | BAKE (BNB Chain) | Caution | attention: mintable |
Every danger or caution any tool gave a long-lived token, with the reasons it listed.
No danger rating on a long-lived token from Unhosted. RugCheck rated 1 of the 6 it covers as danger. GoPlus rated 2 of the 18 it covers as danger. Honeypot.is rated 1 of the 18 it covers as danger. A caution on a token like these is not wrong in itself: it usually names a real power in the contract. It just means a caution alone tells you little, and a danger should be rare.
Test 2: what they say about 98 brand-new tokens
This is when people check: in the first hours, before buying. The new tokens split by chain as Solana 40, Ethereum 21, Base 22, BNB Chain 15. Solana hit our cap of 40 busiest tokens; the other chains had fewer new tokens busy enough to make the cut, and all of them are in.
| Tool | Chains | Tokens | Danger | Caution | OK | No answer |
|---|---|---|---|---|---|---|
| Unhosted | Solana, Ethereum, Base, BNB Chain | 98 | 26 | 33 | 39 | 0 |
| RugCheck | Solana | 40 | 10 | 17 | 13 | 0 |
| GoPlus | Ethereum, Base, BNB Chain | 58 | 12 | 19 | 27 | 0 |
| Honeypot.is | Ethereum, Base, BNB Chain | 58 | 3 | 0 | 11 | 44 |
Two tools looking at the same new token often gave different ratings, and sometimes opposite ones. On the 14 new EVM tokens all three EVM checkers answered, they never all said danger at once.
| Tools | New tokens both answered | Same rating | Opposite (danger vs OK) |
|---|---|---|---|
| Unhosted and GoPlus | 58 | 32 (55%) | 6 |
| Unhosted and RugCheck | 40 | 21 (53%) | 6 |
| Unhosted and Honeypot.is | 14 | 8 (57%) | 3 |
| GoPlus and Honeypot.is | 14 | 11 (79%) | 3 |
“No answer” matters on brand-new tokens: Honeypot.is had none for 44 of 58, mostly because it found no trading pool it supports. A checker that says nothing about a token launched an hour ago has not cleared it.
What each tool flagged most
| Tool | Most common reasons on new tokens it rated danger or caution |
|---|---|
| Unhosted | Much of the trading comes from a farm of related wallets (29) · The launch was bundled: many wallets bought in the same block (21) · The trading looks automated (17) · A ring of wallets trading with each other drives the volume (14) · The creator has rugged other tokens before (7) |
| RugCheck | Low Liquidity (20) · Single holder ownership (6) · Top 10 holders high ownership (4) · High ownership (3) · High market cap per holder (3) |
| GoPlus | sell tax unknown (22) · buy tax unknown (20) · closed source (10) · trading can be paused (9) · mintable (6) |
| Honeypot.is | honeypot (2) · high fail rate (2) · closed source (2) · all snipers honeypot (1) · low fail rate (1) |
In each tool’s own words, with the number of new tokens where it gave that reason.
Same token, opposite answers
| Token | Chain | Age | Unhosted | RugCheck | GoPlus | Honeypot.is |
|---|---|---|---|---|---|---|
| Pokecity 4wop…pump | Solana | 23h | Danger | OK | — | — |
| MONEYBIZ EHPu…pump | Solana | 33h | Danger | OK | — | — |
| SAYLOR3 5MQh…v3JT | Solana | 2h | OK | Danger | — | — |
| KIMCHIBULL GNE7…pump | Solana | 3h | Danger | OK | — | — |
| DARKCAT fpam…pump | Solana | 36h | Danger | OK | — | — |
| TRUMP3 6s2u…Fiue | Solana | 25h | OK | Danger | — | — |
| Trump 0xd6…ecc0 | BNB Chain | 36h | Danger | — | OK | No answer |
| JPMD 0xa6…5253 | Ethereum | 44h | Danger | — | Danger | OK |
New tokens where one tool said danger and another said OK. “—” = the tool does not cover that chain.
Test 3: which warnings came true?
We are tracking the 86 new tokens from this test that still had at least $1,000 in their pool at the check (the rest were already close to empty). On 22 October 2026 (24 hours after the check) and on 28 October 2026 (7 days) we record which pools lost 90% or more of their liquidity, which tokens lost 90% of their price, and set that against what each tool said on day one. The results go here and into the dataset.
The four tools side by side
| Unhosted | RugCheck | GoPlus | Honeypot.is | |
|---|---|---|---|---|
| Chains in this test | Solana, Ethereum, Base, BNB Chain | Solana | Ethereum, Base, BNB Chain | Ethereum, Base, BNB Chain |
| Chains it covers | 13, Robinhood Chain included | Solana (and Fogo) | 44 on its check page, Robinhood Chain included; no Solana | Ethereum, BNB Chain and Base only |
| Can-you-sell test | Stored test trade on EVM chains | — (reads Solana token settings) | Honeypot and tax detection | Live buy and sell simulation, plus test sells from real holders |
| Owner powers | Solana: mint, freeze, fee and hook settings; EVM: the scam rating’s reasons | Mint and freeze authority, mutable metadata | Mint, proxy, blacklist, pause, hidden owner, balance changes | Proxy and open-source flags |
| Holders | Top 10 share, launch snipers, bundlers and insiders still holding | Top holders, insider networks | Top 10 holders, LP holders | Holder count |
| Liquidity | Pool size, market cap vs liquidity | Locked share of liquidity | LP holders and locks | Pool reserves |
| Answer | A rating with a one-line reason | GOOD / WARNING / DANGER | Risky and attention items | A risk level and the taxes |
| Free, no sign-up | Yes | Yes | Yes | Yes |
Coverage and features as shown on each tool’s public pages and API responses on 21–22 October 2026.
Two tools that are often listed as rug checkers do something different and work best next to one of the four: Bubblemaps draws the wallets behind a token, which shows when one team hides behind many holders, and DEX Screener shows the chart and the pool, not a verdict.
How to check a token for a rug in a minute
- Use the contract address Copy it from the project’s own account or the pool, never search by name: copycat tokens share names and logos.
- Can you sell, and at what tax? On Ethereum, Base and BNB Chain, a failed test sell or a sell tax above about 10% is a stop.
- What can the owner still do? Mint new supply, freeze holders, blacklist wallets, pause trading or change balances. On Solana, read the mint and freeze authority.
- Who holds it? The top 10 wallets’ share, and launch snipers, bundlers or insiders who still hold and can sell into you.
- Is there liquidity to sell into? Under about $10k, or a market cap 100 times the pool, means the price on screen is not a price you can sell at. Check whether the liquidity is locked.
When two checkers disagree, go with the worse rating. You can run most of these questions in one go in our free rug checker, then read the token’s AI report for who is buying and selling it.
Questions people ask
What is the best free rug checker?
There is no single best one, because each covers different chains and checks different things. On Solana, RugCheck reads the token’s mint and freeze settings, holder concentration and locked liquidity. On Ethereum, Base and BNB Chain, Honeypot.is and GoPlus test whether a token can be sold and what it taxes. Unhosted covers all four chains in one place and adds the launch snipers, bundlers and insiders still holding. In our test the tools often disagreed on the same new token, so checking two is the safest habit.
What is the best rug checker for Solana?
RugCheck is built for Solana and shows the most Solana-specific detail: mint and freeze authority, metadata changes, top holders, insider networks and how much liquidity is locked. Unhosted reads the same token settings, adds a scam rating and the launch wallets still holding, and also covers EVM chains. GoPlus’s public check page does not list Solana, and Honeypot.is is for EVM chains.
What is the best rug checker for Base and BNB Chain?
Use a checker that test-sells the token, because honeypots are the most common trap on EVM chains. Honeypot.is simulates a buy and a sell, and sells from real holders’ wallets, and reports the taxes, but in our test it found a pool for only 14 of 58 brand-new tokens. GoPlus reads the contract for honeypot code, taxes, blacklists and owner powers and answered all 58; Unhosted shows a stored test trade, a scam rating and the launch wallets and also answered all 58. All three cover Ethereum, Base and BNB Chain.
Can a rug checker tell me a token is safe?
No. A clean result only means no red flag was found when the check ran. An owner can change taxes, liquidity can be pulled in one transaction, and a brand-new token may not have been tested yet. Treat a checker as a way to find reasons not to buy, never as a guarantee.
Why do rug checkers give different answers for the same token?
They look at different evidence and set different thresholds. A simulator flags a token whose test sell fails; a contract scanner flags owner powers such as minting or blacklists even when nobody has used them; a holder check flags a few wallets owning most of the supply. A token can pass one and fail another, which is why the worse of two ratings is the better guide.
How do I check if a token is a rug pull before buying?
Paste the contract address, never the name, into a checker. Then answer five questions: can you sell it and at what tax, can the owner mint, freeze, blacklist or change balances, do a few wallets or launch insiders hold most of the supply, is there enough liquidity to sell into, and is that liquidity locked. Any clear no is a reason to walk away.
The dataset
Ratings are what each tool said at the moment of the check. Tools update their data, so a re-check today can differ; the reasons each tool gave are in the JSON, so any rating here can be traced to its source.