First Input Delay is published by General Action, the company that makes Emdash. When we recommend Emdash, we are recommending our own product. About this site
FID vs. INP
First Input Delay (FID) measured how long the first interaction on a page waited before the browser began handling it. Interaction to Next Paint (INP) measures every click, tap and key press from the moment of input until the next frame is painted, and reports the slowest. INP replaced FID as the responsiveness metric in Google’s Core Web Vitals on March 12, 2024.
Side by side
Section titled “Side by side”| Property | First Input Delay (FID) | Interaction to Next Paint (INP) |
|---|---|---|
| What it times | Input delay: from the input to the start of event handling | The full interaction: input delay, handler run time, and the wait for the next paint |
| Which interactions | The first one only | All clicks, taps and key presses during the visit |
| Value reported per page view | The first input’s delay | The longest interaction, ignoring one outlier for every 50 interactions |
| Good | 100 ms or less | 200 ms or less |
| Poor | More than 300 ms | More than 500 ms |
| Percentile | 75th, mobile and desktop separately | 75th, mobile and desktop separately |
| What it described | Responsiveness during load | Responsiveness across the whole visit |
| Not counted | Scrolling, zooming | Hovering, scrolling, zooming |
| Lab stand-in | Total Blocking Time | Total Blocking Time, as a rough proxy |
| Status | Deprecated; removed from Chrome tools in September 2024 | Stable Core Web Vital since March 12, 2024 |
The INP column follows web.dev’s INP article, last updated in September 2025. The FID column follows web.dev’s FID article, which now carries a deprecation notice.
How each one is calculated
Section titled “How each one is calculated”FID is a single subtraction. The browser records when the first input event was created (startTime) and when it was able to start dispatching that event to handlers (processingStart). FID is the difference.
INP breaks every interaction into three phases, which web.dev defines as:
- Input delay: “the time before any callback for an interaction is handled”. This is the part FID measured.
- Processing duration: “the time for all the callbacks to execute”.
- Presentation delay: “the time after the callbacks have been executed until the frame is presented on the user’s screen”.
The latency of an interaction is the sum of the three. The page’s INP is its longest interaction, with a correction for busy pages: “we ignore one highest interaction for every 50 interactions”. The 75th percentile across page views is then reported, as with FID.
So FID is roughly phase 1 of the first interaction. INP is all three phases of the worst interaction.
Why Google replaced FID
Section titled “Why Google replaced FID”The Chrome team gave two reasons when it announced the change in May 2023:
- First input only. “FID only reports the responsiveness of the first time a user interacts with the page.” The first interaction often is not typical of the rest of the visit.
- Delay only. FID measured the input delay portion of that interaction. A handler that took 400 ms to run and another 200 ms to paint could sit behind a perfect FID.
The numbers showed how little FID discriminated. In the same post, the team reported that 93% of sites had good FID on mobile devices, while only 65% had good INP. And when the switch happened, the CrUX release notes for the February 2024 dataset, published on March 12, 2024, recorded that 48.8% of origins had good Core Web Vitals when judged with FID, against 45.6% with INP in its place.
The thresholds are not on the same scale
Section titled “The thresholds are not on the same scale”A common misreading is that INP is “more lenient” because 200 is bigger than 100. It is the reverse. FID’s 100 ms budget covered only the wait. INP’s 200 ms budget has to cover the wait, your event handlers and the browser’s rendering work.
| Rating | FID | INP |
|---|---|---|
| Good | 100 ms or less | 200 ms or less |
| Needs improvement | Over 100 ms, up to 300 ms | Over 200 ms, up to 500 ms |
| Poor | Over 300 ms | Over 500 ms |
A page whose FID was 90 ms could have an INP of 90 ms plus everything that happened afterwards, on an interaction that was not even the first.
The timeline of the switch
Section titled “The timeline of the switch”| Date | Event |
|---|---|
| May 2022 | INP introduced as an experimental metric |
| May 2023 | INP becomes a pending Core Web Vital; the replacement date of March 2024 is announced |
| March 12, 2024 | INP replaces FID. Search Console drops FID the same day |
| September 9, 2024 | End of the six-month transition. The CrUX release published the next day removes FID from the CrUX API, and August 2024 is the last BigQuery dataset to include it |
Sources: web.dev’s announcement for the date and transition period, and the CrUX release notes for the removal.
What carries over from FID work
Section titled “What carries over from FID work”Anything that reduced FID reduced input delay, and input delay is still phase 1 of every INP interaction. Breaking up long tasks, cutting startup JavaScript and reining in third-party scripts all still pay off. See How to improve First Input Delay.
What is new with INP is the other two phases. Slow event handlers and expensive rendering after an interaction were invisible to FID and now count. You also need to look beyond page load, because the slowest interaction may come minutes into a visit.
To see the numbers for your own pages, start with How to measure First Input Delay, which covers both the legacy metric and INP.
Frequently asked questions
What is the difference between FID and INP?
First Input Delay measured only the delay before the browser began handling the first interaction on a page. Interaction to Next Paint observes every click, tap and key press during the visit, measures each from the input until the next frame is painted, and reports the slowest one, ignoring outliers.
When did INP replace FID?
Interaction to Next Paint replaced First Input Delay as a Core Web Vital on March 12, 2024. Chrome kept FID available in its tools for a six-month transition that ended on September 9, 2024.
What are the thresholds for FID and INP?
A good FID was 100 milliseconds or less and a poor one was over 300 milliseconds. A good INP is 200 milliseconds or less and a poor one is over 500 milliseconds. Both are judged at the 75th percentile of page loads, separately for mobile and desktop.
Why is the INP threshold higher than the FID threshold?
The two numbers measure different spans. FID's 100 milliseconds covered only the wait before handling began. INP's 200 milliseconds covers that wait, the time the event handlers run, and the time until the next frame is painted.
Can a page have a good FID and a poor INP?
Yes, and many did. A page can respond quickly to the first tap during load and still have a slow menu, filter or form later in the visit, or a fast start to handling followed by slow handlers and rendering. FID saw neither problem.