UNUSUAL FOOD. REAL CURIOSITY. Take a closer look.
weirdeats.Find your next bite

YOUR TASTES. VISIBLE CLUES.

A little more
like you.

Explore flavor and texture directions, then see exactly why each food appears. Add selected tasting feedback if you want a more personal starting point.

Your answers and selected feedback stay in this open page, out of URLs, storage and measurement. Reading your journal is optional and explicit. No food is cleared for allergies or suitability by a suggestion.

Use the simple chooserBrowse outside suggestions

What draws you in?

Start with a good bite.

Optionally read your device journal, then select up to 20 exact catalog foods rated 4 or 5. Nothing is selected automatically. Only selected food identities, enjoyment ratings and exact supported flavor/texture tags inform directions; written notes, dates and goals are not used for ranking. Box assortments are omitted from this recommendation set while their contents and terms remain uncertain.

Your journal has not been read.

How this compares with the simple chooser

No learned model or commission-based ranking is used. Listing wording counts two cue points; editorial routes and historical author perceptions count one. Only the strongest evidence for each requested trait counts. Selected enjoyment feedback contributes one vote per distinct food for an exact supported tag, capped at three votes per trait. Enjoyment ratings 4 and 5 are eligible; a dislike does not automatically mean dislike of every tag. Personal words outside these controlled directions are not interpreted.

A selected manual flavor or texture replaces that dimension’s feedback weights. Requested dimensions without a cue stay unknown and earn no points. A preferred familiar/different format adds one small point; it cannot make a missing sensory trait known. Ties use cue points, food name then stable identity. Price, retailer, commission and nationality add no points.

The comparison applies the same quote, country, label and selected-food exclusions to both lists. The reference uses the original chooser’s listing/editorial rules with the strongest feedback direction per dimension (alphabetical tag on a tie), or your manual direction. It uses no historical opinions or novelty nudge. Comparing the lists is not proof either is better; reader usefulness is unmeasured. Keep the simple chooser if it serves you better.

Optional · no preferences selected

Keep the label close.

These filters use recorded statements, not a suitability assessment. No food is declared safe for an allergy. An absent statement, empty list or unreviewed label never establishes absence. Recognized allergen names are matched to the recorded lists; unrecognized wording stays unknown. Inspect the full statement.

Exclude records that declare these in contains or may-contain statements
Other recorded statements

If you exclude unknowns, only records with a matching stated claim or sourced zero amount can remain. The current catalog’s ingredient labels are untranscribed, so many choices will produce no results. May-contain and shared-facility statements remain on each source record; a free-from or vegan claim does not clear cross-contact.

Choices stay out of shared URLs. Leave this unchecked for this page visit. No analytics are enabled.

Reasons before the next bite.

Public evidence loads only when you ask.