FSRS is the newer spaced repetition algorithm in Anki. Here is what changes in practice, when it matters, and when it does not matter much.
FSRS is an algorithm that tries to show you a card shortly before you would forget it. Compared with Anki's older SM-2 system, it schedules hard cards more intelligently and cuts some wasted reviews. In practice, that often means fewer cards in each session while holding about the same retention, especially once you have enough review history for the scheduler to learn from.
| Feature | FSRS | SM-2 (older Anki default) |
|---|---|---|
| Forgetting prediction | Memory state model (stability + difficulty) | Interval multiplier formula |
| Hard cards | Shorter intervals, closer tracking | Can over-schedule, leading to failures |
| Calibration | Learns from your review history | Fixed formula, no personal calibration |
Page design based on study habits around PDFs, papers, and class materials.
SM-2, the older default, mainly stretches or shrinks intervals based on how you rated the card. FSRS models the card more directly by asking how stable the memory seems and how difficult that card is for you. The result is fewer unnecessary reviews on cards you already know and closer tracking of the ones that keep slipping away.
If your deck is easy and you almost never miss cards, the difference can feel small. It matters more when the material is uneven and a handful of cards keep tripping you up.
Switch if you already review consistently, your deck has real history behind it, and you want fewer wasted reviews without lowering retention. That is where FSRS usually feels most worthwhile. It is especially useful when some cards are easy, some are stubborn, and the older scheduler treats too many of them in roughly the same way.
Do not overthink the switch if you are still learning how to write good cards or if the habit itself is not stable yet. In that stage, card quality and consistency matter more than the scheduler. FSRS helps most after the basic review loop already exists.
FSRS is built into Anki from version 23.10 onward. Open deck options, find the spaced repetition settings, and switch the algorithm to FSRS. If you already have review history, you can also run the optimizer so the schedule reflects your own data instead of relying only on the default values.
For most users, yes. It predicts forgetting more accurately and tends to reduce total review time while holding about the same retention rate.
Go to deck options, find the spaced repetition algorithm setting, and select FSRS. Available from Anki version 23.10 onward.
Run it once you already have meaningful review history. The optimizer works best when it has enough past answers to fit the schedule to your own deck instead of relying only on defaults.
No. FSRS works from your existing review history and gets more accurate as it gathers more data.
Some alternatives, like SocriFlow, use FSRS-based scheduling without requiring the full Anki setup process.
Not at the start. The algorithm matters less than learning to write good cards and keeping a review habit. Switch once you have a deck you reach for regularly.
