The Amazon 75 character title limit: rules, dates and the new Item Highlights field

Short answer: as of 27 July 2026 Amazon caps product titles in all non-media categories at 75 characters including spaces — announced in Seller Central, reported in June 2026. Titles above the cap are gradually rewritten by Amazon's own AI; brand-registered sellers get a 14-day window to review the proposed rewrite. A companion field arrives with it: Item Highlights, 125 searchable characters shown beneath the title.

If you have been using the full 200 characters, the job is not trimming — it is deciding what a title is now for. This guide states the rules with their dates, explains the risk of doing nothing, and gives an order of operations that gets a long title down to 75 characters without dropping what earns the click. Current as of August 2026.

The rule, with dates

• In force since 27 July 2026: a maximum of 75 characters including spaces in the product title.
• It applies to every category except the media categories, which remain exempt.
• The Saudi Arabia, Egypt, Türkiye and UAE marketplaces are exempt as well.
• The source is an Amazon Seller Central announcement, reported in June 2026.
• For comparison: the previous guidance allowed up to 200 characters, while the recommendations had already settled at around 80. What is new is less the short length than the fact that it is now enforced.

Because the exemption list names only four marketplaces, the European and North American marketplaces are covered by the cap. Whether your category counts as a media category is worth checking once in Seller Central against one of your own ASINs: that distinction decides whether you have to act at all.

What happens if you do nothing

27 July 2026 was not a cut-off date on which long titles were abruptly truncated. Amazon is working through this gradually: non-compliant titles are rewritten over time by Amazon's own AI. Brand-registered sellers get a 14-day review window first, in which the proposal can be inspected and replaced.

That is more accommodating than it might have been — and still the worse outcome, for three reasons:

• Brand voice: a model that shortens works with what the title already contains. It does not know your positioning, and it does not know which of five attributes actually separates you from the offer next to yours.
• Keywords: a term that drops out of the title is not automatically lost — but it is not automatically rehoused either. Cut it yourself and you decide whether it moves into Item Highlights or into backend keywords.
• Attention: 14 days is a deadline, not a reminder. When proposals arrive spread across a large catalogue, few teams will realistically review them one by one.

The review window is tied to Brand Registry. Whether and how sellers without it are notified in advance does not follow from the announcement — assume no window if you are unsure.

From 200 characters to 75: the order to cut in

The usual mistake is cutting evenly: one word less everywhere until the number fits. What is left says nothing clearly. A fixed priority order works better — whatever decides the click stays, and everything else moves to another field rather than to the bin.

• 1. Brand. It leads and it stays — it is the one piece of information nobody else can supply for you.
• 2. Product type. The noun people actually search for, in the language of the marketplace, not your internal range jargon.
• 3. One or two attributes that decide the click. Size, quantity, material or compatibility — whichever the purchase in your category genuinely hangs on.
• 4. Everything else. Secondary attributes, use cases, synonyms and spelling variants: out of the title, into Item Highlights or the backend keyword field.

An example with an invented brand. Before, 187 characters in the old format: “Examplebrand Frying Pan 28 cm Stainless Steel Induction Compatible Non-Stick Oven Safe Dishwasher Safe Professional Pan for Induction Gas Electric Ceramic PFAS-Free with Detachable Handle”. After, 73 characters: “Examplebrand Stainless Steel Frying Pan 28 cm, induction ready, PFAS-free”.

Nothing the title had to do is missing: brand, product type, size, material and the two attributes the purchase decision rests on in this category are all still there. Oven safe, dishwasher safe, the detachable handle and the list of hob types all matter — just not enough for the one line a shopper reads in the search results. They belong one row further down.

Two checks before you publish. Read the new title out loud: if it sounds like something a person would say, the semantic search layer parses it better too. Then look at it on a phone rather than in the desktop preview — on a small screen even 75 characters can lose their tail.

Item Highlights: 125 searchable characters, not an overflow bin

Item Highlights is the field Amazon introduced alongside the title cap: 125 characters, searchable, displayed beneath the title in search results and on the detail page. It catches part of what no longer fits above.

Because it is searchable, the field does two jobs at once: it is read and it is indexed. That is exactly why a chain of keywords is the wrong thing to put there. What a person skips over does you little good in search either — and 125 characters fill up fast.

• The use case in words: what for, for whom, in which situation.
• The differentiator that no longer fits in the title — not a third synonym for the product type.
• Compatibility and fit (“for …”, “fits …”): the details people search for, and the ones that prevent a return.
• A limitation, where it settles the purchase — “not dishwasher safe”, for instance. Anything that saves a question saves a return.
• Not here: promotional language, superlatives and repeats from the title. A term already sitting above only costs you space down here.

For the example above that comes to 98 of the 125 characters: “Works on induction, gas, electric and ceramic hobs. Oven safe, dishwasher safe, detachable handle.” Two sentences, and not one word duplicated from the title.

How quickly the field becomes available in every category and on every marketplace may vary. Check it against one of your own ASINs before you build a catalogue-wide plan on top of it.

A migration checklist

• Export the catalogue and sort by title length. Everything above 75 characters is the work list; everything above 120 is the urgent part.
• Remove media categories and the Saudi Arabia, Egypt, Türkiye and UAE marketplaces from that list — the cap does not apply there.
• Prioritise by revenue, not alphabetically: highest-revenue ASINs first, and by hand.
• Per title, fix brand, product type and one or two decisive attributes. Keep everything you cut in a file — it is the raw material for Item Highlights and backend keywords.
• Fill Item Highlights before you shorten the title, or the terms are missing from the listing entirely in the meantime.
• Update backend keywords: whatever dropped out of the title and did not land in Item Highlights belongs in the Generic Keyword field — 500 bytes, and accented characters count as two.
• Proof every new title in the mobile app, not in the desktop preview.
• Note which titles Amazon has already rewritten itself. Review those ASINs first, because they are carrying somebody else's copy.
• After two to four weeks, hold click-through and conversion on the reworked ASINs against the untouched ones — and revisit the titles that lost.

Twenty titles by hand, two thousand not

For twenty ASINs this is an afternoon: open the title, cut, rebuild, proof it twice. But the effort per title is the same whether you have twenty or two thousand — and that is where the arithmetic breaks. Past a certain catalogue depth the problem is no longer the individual title; it is knowing which two attributes decide the click in a given category, and whether there is evidence for them.

That part can be handed off. Torch Listings rebuilds titles and bullets from reviews, competitor listings and demand data, cites every claim to its source, and flags what it cannot verify instead of inventing it — which matters more than usual during a cut, because compressing copy is the fastest way to assert something the spec sheet never said.

What no tool takes over: deciding which attribute your product can actually carry — and looking at the numbers two weeks later.