Choose a LinkedIn recap tool by its data, period and actual output

Write a four-part decision contract

Start with the question the recap must answer. “Which public posts attracted visible responses?” and “How many impressions did my content receive?” require different sources. Then specify exact dates, needed output and acceptable access. A downloadable GIF is not a substitute for a private analytics workbook, even if both are called a Wrapped.

This comparison was checked against public primary pages on October 31, 2026. We operate Wrapped.li, so this is not an independent product review. We did not connect a LinkedIn account, buy a competing plan, import a real workbook or test each provider’s private output. The table separates documented inputs and outputs from the next checks that remain yours.

Compare the six routes on the task, not a universal ranking

MagicPost documents a free recap with its own account and a LinkedIn connection including impressions/follower permissions. Kliver documents a connected retrospective with PNG poster export, plus a separate branded dashboard-report workflow. Oiti’s public recap page describes an English AggregateAnalytics XLSX input and claims browser-local processing. Those are different acquisition paths; this review did not audit their permissions, file handling or private results.

Cleve’s checked page still advertises the 2025 edition and says it fetches public posts from a profile link. Wrapped.li uses public activity for a free preview and paid selected-edition story/exports, without private impression or follower-growth history. LinkedIn’s official FAQ describes its own app-only 2025 activity experience. These observed pages do not prove a particular account’s access to a future seasonal edition.

Keep the period and the product surface visible

A rolling analytics file, a calendar-year poster and a current-year-to-date public sample can all be useful, but they answer different questions. Read the actual dates in the source and output. Oiti’s page asks for a 365-day English export and describes limits in its top-post rankings; that does not make the file a complete calendar-year archive. A missing row cannot be reconstructed just because the tool draws a yearly chart.

Kliver’s help says its sidebar entry appears in December and January while the retrospective remains at the same address outside that period. Its poster is not its branded PDF report: the latter has its own date selection and contents. MagicPost’s help also distinguishes Wrapped from the all-year metrics dashboard. Verify the specific surface instead of transferring every advertised dashboard format, history rule or price to the recap.

Check cost and data access as separate commitments

Free preview, free full recap and a paid analytics subscription are not the same offer. MagicPost says its recap requires no subscription; Kliver is a paid per-profile analytics product; Oiti describes the standalone recap as free without sign-up; Cleve advertises a free recap on its checked page. Wrapped.li’s preview is free and the complete selected-edition story/exports are paid. Verify the current price, billing period, tax and download entitlement at the relevant product step; this article does not invent one price for every profile or region.

Not handing over a password is useful, but it does not by itself establish that a service meets your privacy needs. Read the requested permission scope, processing location, retention and removal route before supplying an account or file. “Parsed locally” in this comparison is Oiti’s published claim, not an independent network audit. Conversely, public activity can still contain names or conversations that should not be repeated in a public recap.

Run three complete choices through the worksheet

Case A needs private impression totals but allows neither an account connection nor an analytics export. None of these documented third-party routes is proven to meet that contract. Stop with an unresolved requirement rather than choosing the prettiest public-profile poster. The user may need to inspect native analytics or explicitly revise the acceptable-source condition.

Case B needs private metrics, accepts an English export and does not want a persistent app connection. Oiti’s documented file path is a candidate, conditional on the workbook’s dates, fields and handling meeting the task. Case C needs a public-supporter story as a PDF or GIF, accepts the collection limits and a quoted one-time payment. Wrapped.li is a candidate after checking the selected edition and format. Neither example is a measured customer outcome or a claim of universal preference.

Finish by checking the artifact you will actually use

The editable worksheet records requirement, source, dates, unknowns, output and cost check. Fill “unknown” when a provider page does not resolve something; do not turn an absent feature description into a global claim that the feature is impossible. For a seasonal native experience, use current official account guidance rather than assuming last year’s launch calendar repeats.

After generation, inspect one important number against its actual source, open the required download and keep the period and sample limit beside the result. Check whether a “top” post means impressions, visible responses or another metric. Save the source and chosen artifact before a seasonal page changes. This closes the selection task; it does not prove all the resulting data, reader value or business effects without those separate checks.

Compare source, output and price on the same surface

The refreshed public-source check on 1 November retains the six routes above. MagicPost’s current help describes its 2026 poster, account/connection permissions and free recap separately from its paid year-round dashboard. Kliver’s public pricing is USD 19 per profile monthly or USD 114 billed yearly before tax; its retrospective PNG and selected-period dashboard report remain separate artifacts. A dashboard PDF label is not evidence of a PDF version of the year poster.

Oiti’s page describes an English analytics-file input, a rolling export window and browser-local processing; those are provider claims, not our audit of private data handling. Cleve’s actual public page still says 2025 and public posts. The current official native FAQ describes the historical 2025 app experience. For any later edition, check actual availability, selected dates and output rather than transferring one product’s year, private-data access or download promise to another.

For Wrapped.li, use its actual free-preview versus paid selected-edition/output boundary and the quote shown for the desired product step. Do not replace a missing final price with a fictional number from our evaluator. The local recap example’s USD 12.00 once is deliberately constructed arithmetic, not a live offer from Wrapped.li or any competing provider.

Use the local recap contract before supplying an account or file

The working evaluator below records up to four candidates against the exact owner/input, metric, period/coverage, required output and acceptable processing/access. Every Yes needs your source/date/finding note; an empty note remains unknown. A No blocks the required contract. All declared Yes means ready for a bounded authorized acceptance check, not a vendor endorsement or permission to publish someone else’s information.

Use the price fields for the actual billed unit and interval, including minimum quantities. One-time cost is kept separate from recurring totals. Save the private JSON and reopen it locally to retain the decision. This is a decision-record format, not an analytics-workbook importer or a generator of vendor recaps. The same generic record can be opened from our analytics comparison; its requirement text and evidence retain the actual context.

The constructed example needs private owner impressions through 31 October 2026, a known capture date and a shareable PNG, while permitting a file and declining a persistent account connection. Its file-based option is unknown until evidence is entered; its public-only option cannot satisfy the required private metric. These are deliberately imaginary options. Neither a source label nor an entered Yes proves that a real file was processed or the actual download works.

Resolve three output and coverage failures

D1 — Constructed partial-year case: a source ending on 31 October is labeled “the complete 2026 year.” Outcome: hold that claim or relabel it with the actual start/end and sample boundary. A 2026 product title does not fill November and December. If the source is a rolling 365-day window, inspect its start date as well; neither label automatically equals January–December.

D2 — Constructed wrong-output case: the task requires a shareable PDF of the specific recap, but the documented surface offers PNG and a separate dashboard PDF. Outcome: the required download stays unknown or unmet. Inspect the correct output and period before paying; a PDF elsewhere in the product is not acceptance evidence for this recap.

D3 — Constructed access constraint: private impressions are mandatory, a persistent connection is declined and an authorized English analytics file is available. A documented file-based route is a candidate if its fields, dates and processing fit. Outcome: check the actual accepted file and required artifact; keep missing fields unknown. If no acceptable source can supply the metric, stop instead of calling a public-profile poster equivalent.