The best time to post on Instagram in 2026, from 1,995 posts

The best time to post on Instagram in 2026, going by 1,995 posts from 201 public accounts, is Thursday through Saturday, in the late evening in US Eastern. Posts in those slots picked up about 10% more likes than the same accounts got the rest of the week. That's the whole effect. It's real enough to lean on when you're choosing between two slots, and small enough that a good post on a Monday morning will still beat an average one on a Saturday night.

If you've read the other studies, you've seen Wednesday at noon, Thursday at 9am, 5am local, and 7 to 9am all crowned as the answer. Those pieces rank raw likes by hour, so a few enormous accounts that post at the same time can make that hour look special. I scored every post against its own account's usual likes first, and most of the drama went away.

Is there a best time to post on Instagram?

Yes, but a modest one. The reason every guide names a different hour is that they measure different things. Some count likes, some count total engagement, some count reach. Some convert to the audience's local time and some don't. And nearly all of them compare raw numbers across accounts, which means the hour a few giant accounts happen to post in wins by default.

Score each post against its own account first and the giants stop mattering. What's left is a gentle lean toward the end of the week and the end of the day. That's what this study measures: 1,995 posts, one year of posting, each post divided by its own account's median likes.

The best times to post on Instagram at a glance

The best day of the week to post on Instagram

Saturday, Friday, and Thursday. In that order, though the three are close enough that I'd call them one block.

Pool the two halves of the week and the gap is easier to read. Thursday to Saturday posts landed at 1.09x their account's normal. Monday to Wednesday posts landed at 0.99x. The ratio between them is 1.11, and the 95% interval runs from 1.00 to 1.24, so it clears zero, barely.

Sunday is the odd one out at exactly normal. It also had the fewest posts in the sample, so I wouldn't read much into it either way.

Share of posts that beat their own account's median, by day:

• Monday: 45%, from 312 posts
• Tuesday: 48%, from 345 posts
• Wednesday: 45%, from 340 posts
• Thursday: 48%, from 243 posts
• Friday: 52%, from 272 posts
• Saturday: 49%, from 258 posts
• Sunday: 44%, from 225 posts

Friday is the only day where more than half the posts beat their account's baseline. The rest sit within a few points of the 47% average.

The best time of day to post on Instagram

Late evening. 11pm ET was the top hour at 1.13x, with 10pm and midnight just behind it. The morning and midday stretch from 8am to 1pm is the low point, at just under normal.

The shape is a shallow U. Engagement is a little above normal overnight and in the early morning, dips through the working day, and climbs back after 4pm. Pool 9pm to 1am and compare it with 10am to 2pm: the evening window came in at 1.13x, the midday window at 0.98x, and the ratio between them is 1.16 with a 95% interval from 1.01 to 1.33.

Two things to keep in mind before you move your whole calendar to 11pm. The sample was discovered globally and every timestamp is in US Eastern, so 11pm ET is morning in Asia and the small hours in Europe. Some of that overnight lift is probably other people's daytime. And the smoothing window is five hours wide, so a real one-hour spike would be flattened here.

The worst times and days to post on Instagram

Wednesday around noon. That's the lowest cell in both charts, and it's still only 2% under normal.

There isn't a punishing slot in this data. The worst day and the worst hour both sit within a couple of points of an account's usual post. If your only free slot is Wednesday at lunch, post then. The cost is a rounding error next to the cost of not posting.

Does the best time change with follower count?

I couldn't tell, and I'd rather say that than invent a tier rule. The sample has 71 accounts under 10k followers, 73 between 10k and 100k, 44 between 100k and 1M, and 13 over 1M. Cut the weekday result by tier and the large tier is a dozen accounts, so any pattern there is a few creators' habits.

The TikTok version of this study found that its Saturday dip concentrated in bigger accounts. I went looking for the same thing here and the sample is too thin above 1M to say.

How this compares to the big studies

The studies you've probably already read each name a window, and they don't overlap:

• Buffer, from 9.6 million posts, names Wednesday as the best day and Thursday 9am, Wednesday noon, and Wednesday 6pm as the best slots.
• Sprout Social, from nearly 2 billion engagements across roughly 307,000 profiles, points to Monday 2 to 4pm, Tuesday 1 to 7pm, Wednesday noon to 9pm, and Thursday noon to 2pm.
• Hootsuite, from more than 1 million posts across 118 countries, gives a different morning or afternoon window for each weekday.
• Later, from more than 6 million feed posts, lands on 5am in the audience's local time.
• SocialPilot, from 7 million posts and about 50,000 accounts, says 7 to 9am overall.

Where this study agrees with them: every one finds a modest effect once you look past the headline hour. Where it disagrees: Wednesday is Buffer's best day and my lowest, and none of them put Saturday or the late evening at the top.

The difference is mostly method. Ranking raw likes by hour rewards whatever hour the biggest accounts use. Scoring each post against its own account measures whether a slot helped that account, which is the question you actually have.

How I measured it

The whole crawl ran through two Instagram endpoints on AnyAPI and cost $1.18 in API calls.

1. Find accounts by niche. I ran 40 creator-niche phrases (home cooking, yoga teacher, personal finance, gaming creator, and so on) through instagram.search_profiles, which returns matching public profiles with a handle and follower count. After dropping private and empty profiles, 359 accounts were eligible.
2. Select across follower tiers. A round-robin across four follower tiers picked 280 of them, so the sample wasn't all tiny accounts.
3. Pull posts. instagram.user_posts returns an account's recent posts with a timestamp and like count, plus a cursor for the next page. The newest page is often still collecting likes, so for any account whose first page didn't hold five mature posts I followed the cursor.
4. Apply the cutoffs. Posts had to be published between 2025-09-02 and 2026-08-19 and be at least 14 days old, so the like count had settled. Accounts needed five scorable posts. That left 211 accounts and 2,126 posts.
5. Drop the bad rows. Nine accounts came back with exactly 3 likes on every single post, which is what the endpoint returns when an account hides its like counts. Those 82 posts went, along with 46 posts reporting zero likes and 3 posts from accounts that fell under five. That left 201 accounts and 1,995 posts.
6. Score each post against its own account. Divide each post's likes by that account's median likes, and cap the result at 6x so one viral post counts once. 153 posts hit the cap.
7. Bucket and average. Convert timestamps to US Eastern, group by weekday and by hour, and take the geometric mean of the scores in each bucket. Hourly buckets are smoothed over a five-hour window. Every interval in this post comes from resampling accounts, not posts, 1,000 times, so one prolific account can't narrow it.

One method note, because it changes the answer. The first pass at this analysis used the median of each bucket rather than the geometric mean, and every weekday and every hour came back at exactly 1.000x. That looked like a flat line. It was an artifact: about 9% of posts sit at exactly 1.0 by construction (one post per account is its own median, and the hidden-likes accounts scored 1.0 on every post), so the bucket median landed on that pile whenever a bucket was anywhere near half above baseline. The geometric mean sees the whole distribution, and that's the measure in the charts above.

The limits of this data

• Likes only. Nothing here speaks to reach, saves, shares, comments, or follower growth, and Instagram's own ranking leans on sends and saves more than likes.
• US Eastern for everyone. The sample is global and I didn't map each account to its audience's time zone. The evening lift is partly someone else's morning.
• A convenience sample. Accounts came from 40 niche searches. A different set of niches could shift a result this small.
• Observational. Content quality, format, and promotion all confound this. An account that saves its best posts for Saturday night will make Saturday night look good.
• Small effects, wide intervals. The day and evening gaps each clear zero, but not by much. A second sample could land them a few points either way.
• Current follower counts. An account that grew during the year is labeled by today's size.

How to find your own best time to post on Instagram

The published numbers, including these, are the prior you use until your account has enough history to measure. Once it does, the same recipe runs on one account in an afternoon:

1. Pull your recent posts through instagram.user_posts, and page past the first response so you aren't measuring posts that are still collecting likes.
2. Divide each post's likes by your median likes over the same window. Cap anything over 6x.
3. Group by weekday and by hour in your audience's time zone, and take the average of each group.
4. Ignore any group with fewer than about ten posts. Thin buckets invent peaks.
5. Give the slot you pick a few weeks before you trust it, then measure again.

If you'd rather not write that, most scheduling tools do a rough version of it from your audience's active hours. That's a fine place to start. It just isn't the same as measuring what your posts actually did.

Frequently asked questions

What is the best time to post on Instagram?

In this sample, Thursday through Saturday, between 9pm and 1am US Eastern. Posts in that window got about 10 to 16% more likes than the same accounts' posts on Monday to Wednesday and around midday.

What is the best day to post on Instagram?

Saturday, at 1.11x an account's usual likes, with Friday and Thursday at 1.09x. Monday to Wednesday sat at about 0.99x. Treat the three end-of-week days as one block.

Does posting time matter on Instagram?

A little. The best and worst day are about 13 points apart, and the best and worst hour about 15 points apart, after scoring every post against its own account. That's worth a nudge when you're choosing a default. It isn't worth holding a good post for.

Why do other studies name a different best time?

Different samples, different metrics, and different time-zone handling, and most rank raw engagement by hour, which lets the biggest accounts decide the answer. Buffer says Wednesday, Later says 5am local, SocialPilot says 7 to 9am. Score each post against its own account first and the field flattens out.

Is there a bad time to post on Instagram?

Not really. The lowest slot in this data, Wednesday around noon, is 2% under normal. Posting at a bad time costs you far less than not posting.

The TikTok version of this study uses the same method, and the Instagram endpoints are what you'd use to rerun it on your own account.

Use data responsibly and follow AnyAPI's Acceptable Use Policy.