Overview: Why Solana TPS and Congestion Matter to Traders
If you trade on Solana DEXes, you’ve probably seen both extremes:
- Swaps that confirm in under a second for fractions of a cent.
- Periods where transactions hang, fail, or need multiple retries even though Solana is marketed as a 60,000+ TPS chain.
This article breaks down, with current data, what Solana TPS actually is, how congestion really works in 2024–2025, and what it means for your swaps, limit orders, and bot strategies.
We’ll focus on:
- The difference between theoretical TPS, observed TPS, and user TPS.
- How Solana’s fee and compute model creates congestion hotspots.
- What recent upgrades (QUIC, local fee markets, Firedancer, etc.) have changed.
- Practical settings and habits Solana traders should use during high load.
All claims here are grounded in public reports, research, and network data as of early 2026.
Solana TPS: Marketing Number vs Real-World Throughput
Theoretical vs observed TPS
Solana’s original design target is often quoted as 50,000–65,000 TPS on a 1 Gbps network. That figure comes from early whitepapers and benchmarks under ideal conditions (no complex programs, no disk bottlenecks, etc.). (en.wikipedia.org)
In practice, several independent sources converge on these ranges:
- Typical mainnet throughput: around 1,000–4,000 TPS sustained when including all transaction types. Messari, Genfinity, and other analysts describe Solana as consistently delivering in the low-thousands TPS under normal conditions. (genfinity.io)
- Average TPS in 2025: community and research summaries put Solana’s average daily TPS in 2025 at ~1,100, up roughly a third from 2024. (linkedin.com)
- Peak stress tests: in August 2025, a mainnet stress event briefly pushed Solana above 100,000 TPS, making it the first major chain to cross six figures in production. (cointelegraph.com)
Those numbers include all transactions: user swaps, NFT mints, governance, system instructions, and validator vote transactions.
Why “user TPS” feels lower
From a trader’s perspective, the number that matters is how many user transactions per second actually land successfully, not raw protocol TPS.
Several factors make user TPS feel lower:
- Vote transactions: Validators continuously send vote transactions to maintain consensus. These count toward TPS but don’t represent user activity.
- Failed transactions: During congestion, many transactions are submitted but fail due to insufficient fees, compute limits, or full block space. They still consume resources and appear in TPS metrics. (reddit.com)
- Hotspot programs: When a single program (for example, a memecoin mint or a popular DEX pool) becomes a hotspot, it can saturate the compute budget for specific accounts even if global TPS looks fine.
This is why you can see dashboards showing thousands of TPS while your single swap keeps failing.
How Solana’s Fee and Compute Model Creates Congestion
To understand congestion, you need to understand how Solana prices and schedules work.
Compute units and block limits
Every Solana transaction consumes compute units (CUs). Each block has a maximum compute budget; once that’s filled, additional transactions must wait for the next block or get dropped.
- Historically, Solana’s per-block compute limit has been raised over time. Recent performance reports note targets around 50 million CUs per block, with room to increase as validator hardware improves. (messari.io)
- Programs like Raydium, Meteora, or Jupiter routing can be relatively compute-heavy compared to simple transfers.
When a few high-demand programs dominate compute, they effectively monopolize block space, creating congestion even if global TPS doesn’t look maxed out.
Local fee markets and priority fees
Solana doesn’t just have a flat gas price like Ethereum L1. Instead, it uses:
- Base fee: a minimal fee for including a transaction.
- Priority fee: an extra fee (in microlamports per compute unit) that you can attach to get your transaction scheduled earlier.
- Local fee markets: fees are applied per account / per hotspot, so congestion on one program or account doesn’t necessarily spill over to the entire network.
In practice, this means:
- If you’re trading a hyper-active memecoin pool, you’re competing with other traders on the same set of accounts (the pool, the mint, the AMM program). Priority fees matter a lot here.
- If you’re doing something on a quiet program (for example, a small NFT marketplace or a niche protocol), you might get through easily even when social media is screaming about “Solana congestion.”
Local fee markets and priority fees were introduced and refined over 2023–2025 specifically to address earlier episodes where spam or a single program could degrade the entire network. (okx.com)
Congestion vs Outages: What Actually Happened 2021–2025
Solana’s reputation was shaped by several high-profile outages early on, but the picture has changed materially.
Outage history in brief
Public incident reports and exchange research summarize the pattern like this:
- 2021–2022: multiple multi-hour outages, often triggered by spam, NFT mints, or bugs in consensus and transaction forwarding.
- February 6, 2024: a ~5-hour outage caused by a software bug that required a coordinated validator restart. (reddit.com)
- After that, upgrades to the main validator client and network stack significantly improved stability.
- By early 2025, several analytics and news outlets highlighted that Solana had completed a full year without network downtime, even as activity hit record levels. (cryptoslate.com)
There is some disagreement across secondary sources about minor incidents in late 2024, but the broad trend is clear: major consensus-halting outages became rare after early 2024, even as TPS and usage climbed.
Congestion without downtime
Even with better uptime, traders continued to experience congestion episodes:
- Users reported periods in 2024 where swaps and staking claims were extremely slow or failing, despite the official status page showing the network as “operational.” (reddit.com)
- These events were usually tied to specific hotspots: NFT mints, airdrops, or memecoin frenzies, rather than the entire network being down.
By 2025, performance reports from infrastructure providers and staking analytics show:
- Average block time stabilizing around 0.4 seconds with much lower variance than 2023–2024. (reddit.com)
- Sustained average TPS ~1,100 across the year, with peaks far higher during bursts. (linkedin.com)
For traders, this means that most of the time Solana is fast and stable, but during extreme demand spikes you’re competing in a very crowded local fee market.
Key Upgrades That Changed the Congestion Story
Several concrete engineering changes have shifted how Solana behaves under load.
QUIC and stake-weighted QoS
Solana moved from a simple UDP-based protocol to QUIC for transaction forwarding, combined with stake-weighted quality-of-service (QoS).
- QUIC provides more reliable, congestion-aware transport between clients and validators.
- Stake-weighted QoS gives priority to transactions forwarded by well-behaved, stake-aligned nodes, making it harder for spam to crowd out legitimate traffic.
These changes reduce the chance that a flood of low-fee spam can saturate the network’s ingress pipeline.
Local fee markets and dynamic fees
As mentioned earlier, local fee markets were introduced so that fees rise on congested accounts without punishing the entire network. Over 2023–2025, these mechanisms were tuned to better reflect real demand and reduce pathological cases where one program could freeze everything. (okx.com)
For traders, this is why you’ll often see:
- One memecoin pool where you must pay high priority fees to get in.
- Another pool or protocol that’s still cheap and fast at the same time.
Firedancer and multi-client architecture
Solana’s original validator client (now called Agave) is being joined by Firedancer, an independent implementation built by Jump Crypto.
- Controlled tests have shown Firedancer handling over 1,000,000 TPS in lab conditions. (everstake.one)
- Public performance reports describe Firedancer as being rolled out across mainnet, with the goal of improving both throughput and resilience by avoiding single-client bugs. (cryptonews.net)
For now, Firedancer’s biggest impact is risk reduction (fewer single points of failure). Over time, as more validators adopt it, it should also increase effective block space and reduce congestion pressure.
What Congestion Feels Like for Solana Traders
From a trader’s seat, congestion shows up in a few concrete ways.
1. Swaps and limit orders stuck in “processing”
On Jupiter, Raydium, Meteora, or other DEX UIs, you’ll see:
- Transactions taking multiple seconds instead of sub-second.
- Phantom / Solflare showing “Transaction may have failed” even though it later lands.
- A higher rate of
Blockhash not foundorTransaction was not processed in timeerrors.
This usually means:
- The accounts your trade touches are hot (for example, a memecoin pool with bots sniping every block).
- Your priority fee is too low relative to competing order flow.
2. Higher failure rates for bots and scalping strategies
If you run a trading bot, congestion manifests as:
- A spike in failed or dropped transactions.
- Worse fill quality because your transactions land later than expected.
- Need to rebalance priority fees and retry logic.
Bots that assume “Solana is always cheap and instant” tend to get wrecked during these bursts.
3. Misleading status dashboards
Official network status pages typically show whether the cluster has consensus and is producing blocks.
- During congestion, the network can be fully operational from a consensus perspective while user experience is degraded.
- Social media often calls this an “outage,” but technically it’s severe congestion, not halted block production.
As a trader, you need to distinguish between:
- Real outage: no blocks, everything is stuck. Historically rare since early 2024.
- Hotspot congestion: your specific trade path is jammed, but the rest of the network is fine.
Practical Tips: Trading Through Solana Congestion
Here are concrete, chain-specific tactics to handle congestion as a Solana trader.
1. Use priority fees intelligently
Most modern wallets (Phantom, Solflare, Backpack) and DEX UIs expose priority fee sliders.
- During normal times, a small or even zero priority fee is usually fine.
- During hype events (new memecoin, big airdrop, NFT mint), you often need to manually raise the priority fee.
Practical approach:
- Start with the “recommended” or “high” priority setting in your wallet.
- If transactions are still failing, incrementally increase the priority fee until you see consistent confirmations.
- Avoid blindly maxing it out; you can overpay significantly if you don’t watch actual network conditions.
2. Watch real-time network metrics
Use Solana-specific analytics to gauge whether you’re fighting the network or just a local hotspot:
- Solscan and Solana Beach: show block times, TPS, and recent transactions.
- Helius, GenesysGo, or other RPC providers often publish status dashboards and incident reports.
If:
- Global TPS is high but stable, and block times are near 400 ms → you’re likely in a local hotspot.
- Block times are spiking and TPS is erratic → there may be a broader network issue; consider reducing risk.
3. Adjust your trading style during peak mania
When Solana is in full mania (for example, a high-profile memecoin launch or major NFT drop):
- Reduce leverage and size on time-sensitive strategies; slippage and failed entries are more likely.
- Prefer simpler routes on Jupiter (fewer hops) to reduce compute usage and failure points.
- Consider placing limit orders (where supported) instead of pure market orders, so you’re less exposed to wild price swings if your transaction lands late.
4. Retry smartly, not blindly
If a transaction fails:
- Check the error message in your wallet or explorer.
- If it’s about insufficient priority fee, increase the fee and resend.
- If it’s about account in use or max compute units exceeded, you may be in a hotspot; consider waiting a few blocks or reducing route complexity.
- Avoid hammering the network with rapid retries; that just adds to congestion and can burn fees without improving your odds.
5. Diversify your execution paths
On Solana, you’re not limited to one DEX front-end:
- Try different aggregators (Jupiter, OKX DEX, SolFi, etc.) which may choose different routes or pools.
- For very hot tokens, sometimes a direct pool trade on Raydium or Meteora is more reliable than a multi-hop aggregated route.
If one front-end keeps failing while others succeed, it may be an issue with their routing or RPC, not the entire chain.
How to Tell If Solana’s Congestion Is Getting Better Over Time
As a trader, you don’t need to read every research paper, but a few metrics are worth tracking over months:
- Average TPS: sustained increases in average TPS (not just peaks) suggest the network is handling more real load. 2025’s ~1,100 average TPS was a meaningful jump from 2024. (linkedin.com)
- Block time stability: lower variance around the ~0.4s target means more predictable execution. Recent community reports show variance dropping significantly in 2025. (reddit.com)
- Outage frequency: long stretches without consensus-halting incidents are a strong sign that the underlying client software is maturing. (cryptoslate.com)
- Client diversity: wider adoption of Firedancer and other clients reduces the risk that a single bug can take the network down. (everstake.one)
If these metrics keep improving while your day-to-day trading feels smoother even during hype, that’s the best indication that Solana’s congestion story is moving in the right direction.
Conclusion: What Solana TPS and Congestion Really Mean for You
Solana’s marketing number (tens of thousands of TPS) and its real-world trading experience are not the same thing.
- In production, Solana typically runs at ~1,000–4,000 TPS, with ~1,100 TPS average in 2025 and peaks above 100,000 TPS during stress tests. (cointelegraph.com)
- Congestion today is less about the entire chain failing and more about local hotspots where you’re competing for compute and account access.
- Upgrades like QUIC, local fee markets, and Firedancer have made outages rarer and performance more predictable, but traders still need to manage priority fees, route complexity, and timing.
If you:
- Understand how Solana’s fee and compute model works,
- Monitor real-time network metrics, and
- Adjust priority fees, routes, and risk during peak demand,
you can trade Solana DEXes far more effectively than someone who only sees “TPS” as a single headline number.
Solana’s throughput story is still evolving, but the tools and data available today already give active traders enough visibility to navigate congestion instead of being surprised by it.