Solana Trading Strategies: Bots, Arbitrage, and Key Risks
Solana trading strategies
The most discussed Solana trading strategies are automated bots that snipe newly launched meme coins and arbitrage across Solana DEXs and centralized exchanges. Solana's 400ms block times and sub-cent transaction fees make both approaches viable in ways they often are not on chains like Ethereum. Users still warn that consistent profitability is difficult and that the meme coin space is full of scams.
Meme coin sniping bots target new token launches on pump.fun, buying on momentum and exiting at preset profit targets. Even sophisticated setups struggle because public signals are usually priced in by the time the order fills. Latency decides everything: if signal detection plus transaction landing takes more than 1.5 seconds, the edge is gone, which is why some builders run custom infrastructure to land transactions in under a second.
For arbitrage, users monitor price differences across Jupiter, Raydium, and Orca for around 20 token pairs and execute when the spread exceeds 0.3% after fees. Larger discrepancies exist between Solana DEXs and centralized exchanges that list SOL pairs, since the two venues serve different user bases. Paper trading results often collapse in live conditions because of slippage, failed transactions, and fees.
Key strategies
- Meme coin sniping bots Buy newly launched pump.fun tokens on momentum and exit at profit targets.
- DEX aggregator arbitrage Watch Jupiter, Raydium, and Orca across about 20 pairs; execute when the spread tops 0.3% after fees.
- CEX-DEX arbitrage Trade price gaps between Solana DEXs and centralized exchanges listing SOL pairs, where spreads run larger.
- Latency optimization Keep decision plus landing under 1.5 seconds using custom infrastructure for sub-second transaction landing.
Automated Trading Bots
Arbitrage Strategies
Risks and Considerations
Are you interested in exploring specific tools or programming languages used for building these Solana trading strategies?
Bottom line
Users discussions on Solana trading strategies frequently highlight the use of automated bots for trading meme coins and for arbitrage opportunities due to Solana's high speed and low transaction fees. However, many Users also caution about the difficulty of achieving consistent profitability and the prevalence of scams.
Community answers 25
What others in the community said:
Today I decided to end my 11-month journey of building a Solana meme-coin copy-cat bot.
It’s been a fun ride, with the system going through three or four major architectural redesigns - from a complex setup with separate services for WebSocket-based data streaming and trade execution originally exchanging data via Postgres, then Redis, then shared memory and a ring buffer - to a lean, gRPC-based design with pure in-memory (RAM) processing; plus countless smaller optimizations along the way. I relentlessly tested latency at every step: built custom parsers, offloaded some logic to Rust, as the bot is Python-based, used only the fastest available libraries, benchmarked both external data providers and in-house built functions, and implemented parallel requests and multi-provider order submission for speed and reliability.
I achieved 25–30 ms latency from receiving a signal to getting my transaction signed by a validator and the signature returned back to me, landing in the same slot about 20 % of the time and within one slot about 75 % of the time. But in the end, it doesn’t even matter still doesn’t make me money. So I’ve decided to call it quits. It’s been an awesome project and I’ve learned a ton, but it’s time to touch grass and focus on something more meaningful.
I've been trading on Solana for about a year now, and one thing that consistently impresses me is how the chain's speed opens up arbitrage strategies that would be impossible elsewhere.
Why Solana for Arb?
On Ethereum, arbitrage between DEXs is basically impossible for retail. By the time your transaction confirms, the opportunity is gone, and you've paid $20 in gas for nothing.
On Solana:
- 400ms block times mean opportunities persist longer
- Sub-cent transaction fees mean small edges are profitable
- High throughput means your tx actually gets included
A Simple Strategy That Works
I run a basic DEX aggregator arb:
Monitor prices across Jupiter, Raydium, and Orca for 20 token pairs
When price difference > 0.3% (accounting for fees), execute
Use Jupiter's routing for the actual swap to minimize slippage
Nothing fancy. A basic Python script with some websocket connections.
Results (Last 30 Days)
- Trades: 156
- Win rate: 71%
- Average profit: 0.25%
- Total profit: ~3.2% on deployed capital
Not getting rich, but it's consistent.
CEX-DEX Arb is Even Better
The bigger opportunities are between Solana DEXs and CEXs that list SOL pairs. The user bases are different, so prices diverge more.
I've been using Backpack as my CEX leg since it's Solana-native. Deposits/withdrawals settle in seconds, which is crucial. On other exchanges you're waiting 10-30 minutes for confirmations.
Tips for Getting Started
Start with the most liquid pairs (SOL/USDC, RAY/SOL)
Don't underestimate slippage - always simulate before executing
Keep some capital on both DEX and CEX at all times
Watch for token-specific events (unlocks, airdrops) that create temporary dislocations
Anyone else running arb strategies on Solana? What's working for you?
I've been working on Solana copy trading and sniper bots for the last 3-4 months and I'm at the point where I genuinely don't know if I'm doing something fundamentally wrong or if the edge just isn't there for retail. Looking for honest feedback from anyone who's been down this road. Started in Python like everyone does. Quickly hit the wall on latency and decided to rewrite the whole thing in Rust. That was a journey on its own but it's now a clean async runtime with proper position management, multiple exit strategies.
Infrastructure:
I went all in. Got a bare metal server in Amsterdam in a colocation facility (Terra Suite area), 125GB RAM, on a subnet that's geographically close to the major Solana validators and Jito block engine.I'm running:
- Triton One bare metal RPC (paid tier, dedicated node)
- Yellowstone gRPC (Dragon's Mouth) for account subscriptions — sub-100ms detection on tracked wallet activity
- Helius as a backup RPC for redundancy
- Jito bundle submission for tip-based inclusion
- TPU direct submission
- Submit to all RPCs simultaneously to maximize landing odds
The bot detects a tracked wallet's buy, builds and signs the same transaction, and fires it within the same slot or the very next one. I've measured it. The speed is there.
I've spent weeks hunting wallets. Pulled data from GMGN, Cielo, manually built a database of wallets with 60-80% win rates and consistent PnL. Built a separate cluster tool that maps relay funding networks and detects insider clusters, and a bunch of others doing $50-200k/month. The wall, every single configuration of mine ends up around -3% to -8% PnL with maybe a 25-30% win rate. Meanwhile the wallets I'm copying are still printing 70%+ win rates. I'm leaning toward giving up. I've learned an absurd amount about Solana's runtime, transaction lifecycle, MEV, AMMs. As a learning project it was incredible. As profit-making endeavor it's been a money pit. But before I shut it down — has anyone here actually made copy trading work at retail scale? Is there a piece I'm missing? I know copy trading they use you as el but i just wondered what if am faster than the bots or other copy traders, here are some wallets i have been watching and trying to figure out how they do it, check them on gmgn. EAZNEs6KeeSbYiaNYGRv2SGoSQq2FFcYWTVNsWxHuEvx, HeN5DgMV6hnZhowhPRaniLVm8kPd1F38z5fFGkfDurJf, 2CrjJGp6XDemP6rTEui2tmwYM8AAM6YowVmzZqEiozw3,5p3BhygmWsqrA1xgVLBV1eUMyJGVSmmQ6nWQ4pQpAdjz
Any feedback appreciated, even brutal.
Cheers
For the past several months, I’ve been building a Solana paper-trading bot focused on newly launched meme coins. The goal is not to blindly copy wallets or chase tokens after they have already moved. I’m collecting transaction-level data, measuring things like buyer participation, buy/sell pressure, market activity, migration behavior, and testing hundreds of rule combinations to see whether any of them hold up across repeated scans.
The project has become much larger than I originally expected. A big part of the work has been making sure the data is actually trustworthy, preventing future information from leaking into backtests, reconstructing realistic entries and exits, and building a system that can evaluate every rule without risking real money. I’ve also been documenting the failures, because most ideas that sound good do not survive once they are tested against real market data.
My best-performing rule so far was tested using a simulated 5 SOL wallet with a maximum position size of 0.1 SOL per trade. During one one-hour paper-trading run, it produced approximately 3 SOL in gross profit. That is an early result—not proof of a permanent edge—and I still need significantly more forward testing before drawing strong conclusions. Meme coins are extremely volatile, and paper results can differ dramatically from live execution.
I’m going to keep sharing the actual results, rule performance, system architecture, mistakes, bottlenecks, and lessons from building the bot. I’m posting most of the ongoing updates on X at . Follow me there if you’re interested in Solana, quantitative research, AI-assisted development, trading bots, or seeing whether this experiment ultimately succeeds or fails.
I spent Sunday morning fixing the mistakes my Solana trading bot was making, here’s what I changed and why
Been building and running an automated Solana meme coin sniper for a few months now as part of my FIRE@45 plan. It runs 24/7, scans for new tokens, buys on momentum, and exits via a staircase of profit targets followed by a trailing stop.
This morning I sat down with the logs from the last 48 hours and found three things that were quietly costing me money. Here’s what I found, what I fixed, and where the bot goes from here.
The problem, three silent killers
- The bot was panic selling tokens that were about to moon
The rug detection logic was triggering on tokens where liquidity dipped 12% and price dropped 5%. Sounds reasonable, except on PumpSwap, a single whale sell can move price -10% on thin liquidity without it being a rug at all.
The smoking gun: one token (I’ll call it Normie) got flagged as a rug at -34% price and the bot sold for a small loss. The same token was re-entered 10 minutes later and hit +241% peak. The false exit cost me the loss PLUS the missed gain, a swing of nearly £3 on a £2 stake.
Looking through the logs I found several of these, tokens that dipped hard, triggered the rug filter, then pumped. All had high price drops but moderate liquidity drops. Real rugs have the opposite signature: liquidity collapses first, price follows.
- The staircase wasn’t working on fast movers
The bot has a 6-level profit staircase, take small % of position at each level, activate trailing stop at +100%. The problem: if a token pumped from entry to +100% faster than one scan cycle (30 seconds), a FORCE TRAIL block fired and marked ALL TP levels as already hit without executing any of the partial sells.
Result: zero profits banked on the way up, full position riding with no safety net. Works out fine when the trail executes cleanly. But on a position where the trail fails, you’ve got nothing banked and take the full drawdown.
- Tokens already mid-run were being permanently rejected
The momentum filter drops tokens with h1 gain above 150% as “too far gone.” Makes sense in theory, you don’t want to buy the top. But I watched a token go from +297% h1 (rejected) to +472% h1 over the next 20 minutes. The bot never entered because it was already “too far gone” at +297%.
The problem is the filter can’t distinguish between a token that’s genuinely topping out vs one that’s mid-run with strong continuing momentum. A token at +297% h1 with m5 still climbing at +8% is completely different from a token at +297% h1 with m5 flat or negative.
The fixes
Rug detection v3:
• Soft rug now requires price below -20% (was -5%) before triggering
• Added proportionality check: liquidity drop must be at least 40% the size of the price drop — real rugs have liq leading price, dumps have price leading liq
• Added 45 second confirmation hold on soft rug: re-checks price after one cycle before selling. Real rugs keep falling. Recoveries bounce. This one change alone would have saved the Normie false exit.
• Hard rug (70%+ liq drain) and confirmed rug (20%+ liq) still exit immediately — no delay on the serious stuff
Staircase restructure:
• Reduced TP1 from 40% sell at +15% to 15% sell at +20%, was banking too much too early, leaving crumbs to ride the trail
• New levels: 15/15/10/10/10% across TP1-TP5, trail activates at +100% with \\\\\\\~50% of position still live
• Added catch-up sells to the FORCE TRAIL block — if a token jumps past +100% in one scan cycle, it now executes all missed partial sells instantly before activating the trail. Safety net is always in place.
• TP6 emergency close moved from +150% to +200%, stops cutting moonshots short
Peak chase filter:
• Tokens at h1 150-350% with m5 still above +5% now bypass the peak rejection and get an immediate entry attempt
• Tokens at h1 350%+ still dropped, data shows these are always topping
• Tokens at h1 150-350% with m5 flat or negative still dropped, genuinely topping out
• All other filters (honeypot, liquidity, mcap, age, deployer blacklist) still apply, just the momentum rejection is bypassed when m5 confirms the run is live
What changes in practice
Before: bot exits Normie at -34% price, misses the +241% move. After: 45s hold fires, price recovers above -20%, bot holds through the dip and rides the trail to +241%.
Before: fast mover hits +120% in one scan, FORCE TRAIL fires, zero partials banked, full position on trail with no safety net. After: catch-up sells bank 50% of position across TP1-TP5 before trail activates.
Before: token at +297% h1 with m5 still +8% gets dropped as “too far gone,” runs to +472% with no position. After: immediate entry attempt, trail activates at +100%, exits somewhere in the +80-100% range from entry.
The cons, being honest
The rug detection changes mean the bot now holds through bigger drawdowns before exiting on genuine slow rugs. A token that bleeds -15% price and -18% liq over several cycles will now hold slightly longer. The proportionality check and 45s window should catch most of these but not all.
The peak chase entries are entering mid-run by definition. On tokens that have already pumped 150-350%, you’re not getting the full move, you’re getting the continuation. Win rate on these entries will likely be lower than normal entries. The bet is that the trail stop captures enough of the continuation to be profitable on average.
The staircase changes mean less protection on tokens that peak at +40-60% and dump. Previously TP1 at +15% (40% of position) and TP2 at +30% (25%) would have banked 65% before the dump. Now only 30% is banked by +40%. Breakeven protection at TP1 means you can’t lose on anything that reaches +20%, but you bank less on mid-range wins.
Where it goes from here
The bot is part of a longer FIRE@45 plan running four parallel systems, Solana sniper, an OANDA forex spread betting bot, a Trading 212 ISA, and a Kraken crypto DCA bot. The sniper is the highest risk, highest reward arm.
The sniper runs on a £39 pot right now with £2 Kelly stakes. Small, intentional, the goal is to prove the strategy before scaling. Every patch like today’s is stress-tested against real trade data before going live.
Win rate sitting at 65.6% across 1,150 trades. Working on improving that alongside the P&L per trade.
Happy to answer questions on any of the logic, always more interesting to discuss the why than the what.
Running as part of a build-in-public FIRE journey. Tracking everything.
Hey everyone, been holding SOL for a while but never actually traded memecoins. I keep seeing people posting insane gains and I want to get in but honestly have no clue where to start.
Like how do you even find the good ones before they moon? Do I need to join alpha groups on Discord for example? And how do you not get rugged, I've heard so many horror stories.
Any advice for someone just getting their feet wet? Don't roast me too hard lol
If it was really working then you wouldn't be posting here. Rather mint money in silence
In the early days of Solana when there was only serum, raydium and Orca, I would do very profitable arbitrage. Nowdats it isnt worth it.
arb opp were great in 2021 not now
ngl this is interesting but also feels like classic “overfitting your bot to past trades” problem
you fixed real issues for sure, but markets especially meme coins are messy, so every rule you add also creates a new way to break it
I’ve seen this pattern a lot, you improve edge in one area and accidentally kill performance in another without noticing until live runs change
respect for actually running it with real money though, most people never get past backtests or theory
just make sure you don’t end up spending more time tweaking than letting the system actually breathe and show real stats
To be honest I have tried all the setups you just mentioned in like 2022 2023 2024 none of it worked, the biggest money loss I've got when followed call channels on Telegram bought their exit liquidity TP's many times, got rugged, dumped on me etc... way to frustrating untill then I decided to build my own bot on TG I've used some skills that you can find on Gmgn api's , Alchemy and helius RPC, and being late by 5 seconds is not that bad specially when the token didn’t pumped too much before you buy, the main problem is to distinguish the holders quality if they're bundle or wash or volume bots etc... that's the hard thing to do, my honest thoughts and based on 4 years experience on memecoins the speed isn't the main things to resolve, bcs when I miss a token that pumped by seconds I'm happy to see that my trade didn't trigger due to slippage it would be buying the top and pray to keep pumping more, what you should fix is the data quality and how to get high accurate and strong buy signals based on real and high quality data, I'm not trying to sell you my bot, just check my website I have mentioned all the data that I'm getting to have my own signals, try to use it with your secret sauce, and the Execution bot on tg is completely open source and have features that no bot have, specially the exit strategies that are based on the token score because a 75pts token is not the same as a 99pts so the exit strategy shouldn't be the same, check it worth to copy some ideas enjoy it. Dm on tg if you need any tips would be happy to share knowledge.
I’ve been stuck rebuilding this same project for quite a while I can’t really help per-se since I’m running into different walls currently. I’d be happy to connect and trade info/work to help each other out if you’re interested
Um... nope. I don't think you're profitable. In fact, you're too slow. You're literally detecting, calculating the opportunities, and finally executing the arb offchain. By that time, the opportunity has already gone.
Been trading Solana memecoins for 18+ months now. I've used pretty much every bot at this point, so figured I'd share what I've learned.
Quick context: I'm a volume trader, mostly playing pump.fun launches and riding momentum. Your mileage may vary if you trade differently.
🥇 Axiom - Best for DesktopLink:
What it does well:
- 500ms execution (fast enough for most trades)
- Points system that could be worth a lot if token launches
- Clean UI, easy to navigate
- Y Combinator backed (adds credibility)
- ~68% market share right now
The catch:
- No instant cashback
- Points value is speculative
- Higher fees than some competitors
Best for: If you believe the airdrop will be substantial. You're essentially farming points while trading.
My take: I keep volume here specifically for the airdrop. The points could be worth nothing, or they could be huge. Asymmetric bet.
🥈 Fomo - Best for MobileLink:
What it does well:
- Outstanding mobile experience (apps)
- Social element (follow traders, interact, engage)
- Jumped to #2 Solana volume
The catch:
- Less "deeper" analytics than Axiom
- Subpar desktop experience vs. Axiom
Best for: Traders who are on-the-go and seek convenience.
My take: This is where I'm moving more volume.
What I Actually UseSplit between Axiom and Fomo:
- Axiom: For the airdrop farming and elite desktop experience
- Padre: Social element, mobile-friendly
There's no "best" bot - it depends what you value:
- Want potential airdrop? → Axiom
- Want the social experience? → Fomo
- Want deeper analytics? → Axiom
- New to this? → Fomo
Disclosure: I use referral codes. Links give you fee discounts and support me. DYOR, this isn't financial advice.
I'm working on one for 2 years now. It's a real challenge
I think your biggest problem is decision making speed and execution speed. If decision + landing takes > 1.5s, than your edge is already gone. I'm doing advanced security analysis without relying on any api, I cache and index everything. I'm getting 0-500ms decision time and 100-1000ms landing time. (Zero block execution)
Check my bio, I also built my own solana memecoins ai scanner and trading bot, you might find some useful info to dev yours, I have the Execution bot on TG that is completely open source and free, I hope it could be helpful for you.
Feel free to have a look at a real-time token on just scroll to the end and you will see a bubble map demo, all the data you or your trading bot needs is right there. If you'd like to try this with a token you are currently looking at or thinking of trading just change the mint address at the end of the URL to the mint address of your choice. I'm just looking to help people who have built trading bots using AI to not get rug pulled by people who are there for themselves. The tool will also flag good tokens that you can do quite well from. Any questions, just shoot me a DM and I'll gladly assist. Thanks
Hey all, hoping some experienced folks can point me in the right direction. I've been at this for a while and I'm a bit stuck, so I figured I'd lay it all out honestly and ask for help.
The short version: I built an automated bot that trades brand-new tokens on pump.fun and Meteora. It works, it places real trades, but over hundreds of trades it just slowly bleeds money. Small wins, but the losers add up and it ends up red.
Stuff I've tried (a lot of it):
- Sniping launches — buying tokens the second they're created. Too slow to be first, so I'm usually buying after the initial pop.
- Copytrading — finding wallets that seem to consistently win, then automatically buying whatever they buy. A tiny handful of wallets actually looked good on paper, but most just lose money when copied at scale.
- Following "funders" — tracking the wallets that fund the developers, hoping to get in before the crowd.
- Buying graduations — waiting for a token to "graduate" (hit the milestone where it moves to a bigger exchange) and buying then.
- Pullback/survivor buys — buying tokens that dipped but survived, betting on a bounce.
- Whale / big-buy detection — reacting when a big buyer shows up.
- Order-flow signals — measuring whether buying pressure heavily outweighs selling, and only entering when it does.
- Telegram signal groups — auto-buying tokens that certain callers post.
- Creator reputation — tracking developers with a history of launching runners vs. rug pulls.
- Tons of exit strategies — fixed take-profit, trailing stops, "moonbag" partial exits, laddered sells, etc.
Why I think it's not working (my best guess):
- I'm always a little late. By the time my bot detects a signal and actually fills the buy, the easy move already happened. The winners I copy tend to be up ~1.5x before I even get in; I'm buying the top.
- Reacting to public info is a losing game. Anything I can see (a spike, a whale, a big buy), everyone else can see too — so I'm chasing.
- Volume without selectivity bleeds. The more indiscriminately I trade, the worse it gets. The only pockets of green were a very small set of specific wallets/setups, and even those are thin and hard to trust.
So basically: selection and speed seem to be everything, and I'm losing on both.
I'm not doing this from a laptop clicking buttons — it's a proper automated bot running 24/7 on a server. I use a gRPC stream (Yellowstone) to watch the blockchain in real time, which means I detect new tokens and wallet activity basically instantly (under a second). That part is genuinely fast.
- Entry (buy) latency: usually around 3.5–5 seconds from decision to the trade being submitted, and it's frustratingly inconsistent — about 1 in 4 buys land in under a second, but more than half take 3+ seconds, and bad ones spike to 8–15 seconds.
- Exit (sell) latency: even slower, roughly 9–17 seconds end-to-end.
- The actual on-chain confirmation is fine (~1 second). The delay is all in the quote/RPC path before submitting.
So detection is sub-second, but by the time I actually fill, several seconds have passed — and in this game a few seconds is the difference between the entry and the exit.
My question to the community: For those who've actually made this work — is the edge mostly in being faster (better infrastructure/latency), in finding genuinely predictive signals that aren't public yet, or is this whole space just a coin flip minus fees for retail bots? Am I chasing something that structurally can't work, or is there a category of edge I'm completely missing?
Happy to share more detail. Really appreciate any honest input — even "you're wasting your time" is useful if that's the reality.
Thanks 🙏
Spent the week pulling on-chain data and reading the forensics (Pine Analytics, Bloomberg's cabal reporting, etc.). The uncomfortable conclusion: the leaderboard "geniuses" aren't out-trading anyone. Over half of pump.fun tokens are sniped in the creation block by wallets the deployer funded — they bundle their own buy at the bottom, manufacture a clean-looking distribution across 20+ fresh wallets, let retail and copy-bots chase it up, then dump. 87% of those creation-block snipes are profitable — a win rate that's only possible if you already know the launch is coming, because you're running it. The practical takeaway for anyone building a trading agent: don't try to copy the winners (their wallets are disposable and already exited — you become exit liquidity), and don't chase speed. Filter. The edge a small trader can actually hold is avoiding the launches that are engineered to dump on you. I ended up building that filter into a tool — it gives one verdict per mint: are insiders positioned to dump on you? (bundled launch, single-wallet concentration, shared-funder cluster, serial-rug dev). Free, no signup, and every flag links to the on-chain tx so you can verify instead of trusting a score. Happy to share the link in a comment if useful — mainly posting because the "the leaderboard is mostly insiders" finding genuinely changed how I think about entries.
New to Solana memecoins? Here's the no-BS guide to getting started.
Step 1: Set Up a Wallet
- Download Phantom (browser extension or mobile). This is the standard Solana wallet.
- Create a NEW wallet just for trading. Don't use your main holdings wallet.
Step 2: Get Some SOL
- Buy SOL on Coinbase/Binance/Kraken and send it to your Phantom wallet. Start small - $50-100 to learn.
Step 3: Choose a Trading Bot
Manual trading on DEXs is slow and you'll miss opportunities. Bots give you speed and better tools.
For beginners, I recommend Axiom. It's the most popular trading bot and very reputable, safe + easy
Step 4: Start Small
Seriously. Your first trades should be tiny. Learn how:
- Buy/sell works
- Slippage affects your trades
- Gas fees eat into profits
- 99% of memecoins go to zero
Step 5: Learn to Read Charts
Use DEXScreener or Birdeye to:
- Check token liquidity
- See holder distribution
- Spot red flags (dev holds too much, honeypot, etc.)
The Hard Truth:
Most people lose money trading memecoins. The ones who win:
- Trade with money they can lose
- Take profits (don't get greedy)
- Learn from losses
- Use proper tools
Full beginner's guide with bot setup tutorials: How to Use a Solana Trading Bot
Questions? Ask below. No stupid questions - we all started somewhere.
After months of private testing, Im finally ready to release my Solana AI trading bot to the public.
I started developing it 6 months ago, using machine learning algorithms to analyze market sentiment, order book liquidity, and high-frequency trading patterns
The results? Consistent 30-40% daily returns
But I get it - you dont trust random bots. Neither would I.
Thats why the minimum deposit is only 0.1SOL
No risk - just test it and see for yourself
If you are interesed: dm me because users banned telegram links
Replies (0)
No replies yet. Be the first to reply.