I use V2 when I want a pool I can deposit into and leave alone. I use V3 when I have a reason to manage a price range. That distinction has saved me more money than chasing whichever interface advertises the larger yield.
For a routine swap, start with the route, not the version label. V2 is the sensible default when the pair has dependable liquidity there and the trade is ordinary: buy, sell, move on. The position is broad, execution is easier to reason about, and there is no range to monitor. V3 becomes worthwhile when its quoted price is materially better after fees and price impact, or when the pair’s liquidity is clearly deeper in a usable range.
Do not treat V3’s capital efficiency as free money. Concentrated liquidity only works while the market remains inside your selected range. If price leaves it, the position can stop earning fees and become effectively one-sided. A narrow range may look excellent in a calm screenshot and be useless after one volatile session. For a position I cannot check regularly, I would rather accept less efficiency than turn liquidity provision into an unattended trading strategy.
What I use for liquidity
For passive liquidity, V2 is the less demanding choice: supply both assets at equal dollar value, collect the pool’s trading fees, and accept the usual inventory drift. It is not risk-free, but the operational burden is low.
For active liquidity, V3 is the better instrument. I choose a wide range when I want fewer adjustments and a narrow range only when I have a clear view of where the pair will trade. The relevant question is not “what is the advertised APR?” It is “how often will this position be in range, and what will rebalancing cost me?” Fees vary with volatility, liquidity, concentration and trading volume, so a displayed rate is a starting signal, not a forecast.
When I am checking a live quote, available pools and the relevant swap or liquidity route, I use quickswap as the working interface. I compare the actual output for my size, then check slippage, the network selected and the token contract before signing. A cheap transaction is still a bad trade if the route is thin or the asset is wrong.
My rule is simple: V2 for convenience, V3 for deliberate management, and neither when the quote is poor. I would change that rule if a V3 route consistently delivered better execution for the same pair and size, or if I had automation that could manage the position without turning every price move into another decision.