Prompt Details
Model
Chat - (gpt-5.6)
Token size
724
Example input
[TASTING_NOTE]: Crisp lime zest, green apple, white flowers, and mineral finish.
Example output
{"text":"Crisp lime zest, green apple, white flowers, and mineral finish.","spans":[{"text":"Crisp","label":"texture","start":0,"end":5},{"text":"lime zest","label":"flavor","start":6,"end":15},{"text":"green apple","label":"flavor","start":17,"end":28},{"text":"white flowers","label":"aroma","start":30,"end":43},{"text":"mineral finish","label":"finish","start":49,"end":63}],"label_legend":{"aroma":"smell, nose, bouquet, or aromatic impressions","flavor":"named fruit, spice, plant, mineral, earth, wood, or other flavor impressions","texture":"mouthfeel, body, weight, structure, tannin feel, creaminess, crispness, or similar tactile sensations","taste":"basic taste or palate sensations such as sweet, sour, bitter, salty, acidic, or balanced taste language","finish":"an explicit aftertaste or ending expression, including wording that describes the length or character of the finish"},"needs_review":[]}
By purchasing this prompt, you agree to our terms of service
GPT-5.6
Wine Sensory Span Tagger converts an existing English tasting note into a compact JSON annotation record. It preserves the note exactly, extracts explicit aroma, flavor, texture, taste, and finish phrases, and gives each span zero-based character offsets. Ambiguous cases are surfaced in a review list instead of being silently guessed. Use it for wine catalog tagging, tasting-room search, or small annotation datasets. It does not write reviews, assign scores, or certify sensory quality.
...more
Added 2 weeks ago
