Money velocity metrics are important measures of a currencies underlying value as a medium of exchange. So, knowing the on-chain exchange volume measured in BTC is important for Bitcoin as it is for every cryptocurrency, but I don't see that data anywhere.
I looked for 2025 on-chain BTC transaction volume and failed at finding that data. I was able to find one chart of BTC transaction volume but it included the change to the sending address, which makes the data wildly misleading. It was also only a daily chart without apparent links or methods for annual data.
You’re spot on. Looking at raw on-chain data in 2026 without filtering is like trying to count real traffic by including every car that just pulls out of a driveway and backs right in. The UTXO model inherently inflates 'volume' because of change outputs.
If you're digging for the 2025 stats, here’s the state of the art right now:
Entity-Adjusted Volume is the Standard: Most of us now rely on Glassnode’s 'Entity-Adjusted' metrics or Chainalysis reports. For 2025, the 'clean' volume was significantly lower than the raw total, especially with the massive 'Whale Shadows' we saw late last year when long-term holders moved millions of BTC.
The 2025 Sell-off Data: Reports from early 2026 (like TRM Labs and Bitcoin Magazine Pro) indicate that 2025 saw a record transfer of BTC from early 'whales' to retail and ETFs. That 'velocity' is real economic activity, but it’s often buried under billions of dollars in internal wallet reshuffling.
The 'L2 Leakage': We also have to accept that 'Money Velocity' is increasingly invisible on-chain. With the Lightning Network and other Layer 2s maturing in 2025, a growing percentage of medium-of-exchange volume never touches the base layer. On-chain volume is becoming more of a 'settlement' layer for large chunks rather than a daily spending metric.
Check out the NVT (Network Value to Transactions) Ratio on Glassnode—they specifically use the adjusted volume you're looking for to filter out the noise. It's probably the most honest metric we have for 2025's actual throughput