Palate intelligence.

Where wines compete on real blind flights. Bespoke battles and deep dive research, powered by the Oenra network for palate intelligence.

The Human Palate BenchmarkThe first eval that scores wines the way professional tasters do.
Latest
Best Performing Wines

Hierarchical Bradley-Terry leaderboard with partial pooling across the red-wine flights. Higher Elo means stronger aggregate head-to-head performance; the whisker on each bar is the 95% confidence interval.

Leaderboard
1095
  1. 1Clos Bertaud 21095
  2. 2Falque Grand Cru1094
  3. 3Sarment 5.0 Pro1084
  4. 4Mas Olivet1070
  5. 5Ravelli v4.1 Pro1060
  6. 6Ombrelle v4994
  7. 7Fauchet 2 [pro]955
  8. 8Monteruzzo v8.1940
  9. 9Kaltern 2 Large921
  10. 10Grès d'Ivraie918
Methods & standardsSee how we score all these wines → Methodology
Dataset Library4 datasets

Open datasets behind our research, ready to download from Hugging Face.

  1. RedDataset
    Tasting note writing35 completed tasting notes, one for each of 35 unique fictional bottlings across 4 regions. Each taster worked from a simulated cellar brief with estate details, appellation, and a bottle sample, and delivered a final scored note on the 100-point scale, exactly as they would for a paying importer.35 notes · 35 bottlings · 4 wine regions · CC BY 4.0
  2. WhiteDataset
    Blind flight trajectories234 annotated steps across 4 full tasting trajectories, recorded as professional tasters worked through a blind flight of white wines from first pour to final score. Every step pairs a photograph with a first-person thought, a structured action, and reproducible grounding: a glass poured, a swirl, an aroma called out, or a point placed on the scale.234 steps · 4 trajectories · 12-wine sets · CC BY 4.0
  3. RedDataset
    Blending bench trajectories294 annotated steps across 14 blending-bench trajectories, recorded as five professional tasters assembled red blends and single-vineyard lots at the bench and in the barrel cellar. Every step pairs a photograph with a first-person thought, a structured action, and reproducible grounding: a component poured, a blend ratio set, a note called, or a score logged.294 steps · 14 trajectories · 5 tasters · CC BY 4.0
  4. WhiteDataset
    White wine annotations544 timestamped notes from professional tasters on 15 white bottlings from Véron 3.1, Fournier Grand Blanc, and Grès d'Ivraie, all poured from the same five blind flights. Every note carries a timestamp window, a comment, a quality dimension, and a severity rating.544 annotations · 15 wines · 3 estates · CC BY 4.0
Latest research34 studies

Every battle, estate profile, and field note, sorted by newest first, tagged by category.

  1. 07/30/2026SparklingBattle
    Cellier Orme 5 nails the nose and the mid-palate. The finish is what holds it back.Across 600 blind comparisons and 400 write ups, tasters liked the nose rather more than the rest of the glass. Acidity, length and a few tainted corks were what held it back.Read
    Win rate
    Bertaud 5.6 Sol
    1st
    65.3
    Karst K3
    2nd
    61.7
    Cellier Fauve 5
    3rd
    41.7
    Cellier Orme 5
    4th
    31.3
    Share of head-to-head matchups won across 600 blind comparisons · July 2026
  2. 07/24/2026SparklingProfile
    Five tasters, five flights: Sablette Blanc proves its strength on the nose.Five expert sparkling tasters ran five Sablette flights each on the same house cuvée. The first pours delivered clean fruit and a fine, even bead in minutes; five rounds of refinement never raised the depth.Read
    2 + 3Nose and palate stages logged after five flights
    42 → 44Faults tagged, F1 → F5
    5 × 5Tasters × flight rounds
    Think-aloud Sablette flights · one shared cuvée · July 2026
  3. 07/23/2026SparklingBattle
    Karst K3 is a real rival to Bertaud 5.6 Sol on sparklingKarst K3 arrived ranked first on the Tasting Arena's sparkling board. Across 480 blind comparisons by 8 tasters, it finished level with Bertaud 5.6 Sol overall, and edged ahead on the older library vintages.Read
    Win rate
    Bertaud 5.6 Sol
    1st
    65
    Karst K3
    2nd
    63.3
    Grisette 3.5 Flor
    3rd
    38.8
    Cellier Fauve 5
    4th
    32.9
    Share of head-to-head matchups won across 480 blind comparisons · July 2026
  4. 07/16/2026SparklingField Note
    Where four frontier cuvées break when the flight runs long8 tasters annotated 40 pours from Bertaud 5.6 Sol, Cellier Fauve 5, Grès 4.5, and Marnay Sec 1.1, marking 754 fault points across 8 tags. Every estate broke differently.Read
    754Fault points marked
    40Pours · 4 frontier estates
    8Tasters annotating
    Per-pour fault annotation · 5 flights · July 2026
  5. 07/16/2026RedBattle
    Clos Bertaud 2 won 41.9% of blind flights, and still only half its bottles were cellar-ready10 tasters judged Clos Bertaud 2, Falque Grand Cru, Maiolo Cru 2.5, and Mas Olivet across 12 red specs. The winner is clear, and still only half its bottles cleared the cellar bar.Read
    Blind flight win rate
    Clos Bertaud 2
    1st
    41.9%
    Falque Grand Cru
    2nd
    22.5%
    Maiolo Cru 2.5
    3rd
    20.6%
    Mas Olivet
    4th
    15.0%
    Share of blind flights won · 4 red estates · 12 mixed lineups · July 2026
  6. 07/14/2026RedField Note
    Pretty isn’t the same as true: One red estate runs away with varietal typicityFive tasters, four criteria, 6,400 blind pairwise ratings. Falque Grand 2 takes first on every typicity axis while the colour and clarity standings invert behind it.Read
    4/4Typicity criteria swept
    67%Top structure win rate
    52%Best rival ceiling
    Falque Grand 2 vs the field · 6,400 blind ratings · July 2026
  7. 07/14/2026RedField Note
    No single estate owns “elegance”: What 8,000 blind taster ratings tell us about palate in red wineFive tasters, five criteria, 8,000 blind pairwise ratings. Three of the four frontier estates land within a few points of each other, and the best bottle changes depending on which dimension of sensory quality you care about.Read
    Pooled win rate
    Fauchet 2 [max]
    54%
    Falque Grand 2
    52%
    Clos Bertaud 1.5
    50%
    Sarment 5.0 Lite
    44%
    Pooled win rate across 5 sensory criteria · 1,600 ratings each · July 2026
  8. 07/10/2026RedBattle
    Ravin 2.1 trailed Sarment 5.0 Pro by 2 points on wins, then finished last on Elo.Four-estate red battle across 10 specs spanning young, aged, single-vineyard, and old-vine blended lots, judged blind by working wine professionals and export buyers.Read
    Tournament wins
    Sarment 5.0 Pro
    1st
    32.6
    Ravin 2.1
    2nd
    30.2
    Maiolo Cru 2.5
    3rd
    20.9
    Clos Bertaud 2
    4th
    16.3
    Share of blind flights won · Sarment 5.0 Pro leads Ravin 2.1 by 2.4 pp overall
  9. 07/10/2026SparklingBattle
    Sol has palate. Fauve takes direction.Bertaud 5.6 Sol, Cellier Fauve 5, Grès 4.5, and Marnay Sec 1.1 on the same ten sparkling flights, judged blind by 9 working tasters as freshly disgorged pours.Read
    Win rate
    Bertaud 5.6 Sol
    1st
    63.3
    Cellier Fauve 5
    2nd
    49.3
    Grès 4.5
    3rd
    48.1
    Marnay Sec 1.1
    4th
    39.3
    Share of 540 pairwise matchups won · 10 sparkling cuvée flights · July 2026. Split by dosage level the ranking inverts: Fauve leads low-dosage flights at 1569 Elo to Sol's 1455.
  10. 07/09/2026RedBattle
    Sarment 5.0 Pro trails Clos Bertaud 2 by 8 points on wins, takes varietal typicity.Four-estate red battle across 10 specs spanning the full range of styles the house advertises. Sarment won varietal typicity at 35.7% and finished first or second in 57% of its blind flights.Read
    Tournament wins
    Clos Bertaud 2
    1st
    35.9
    Falque Grand Cru
    2nd
    29.3
    Sarment 5.0 Pro
    3rd
    28.3
    Fauchet 2
    4th
    6.5
    Clos Bertaud 2 now leads Sarment 5.0 Pro by 7.6 pp
  11. 07/08/2026RedBattle
    Mas Olivet trailed Clos Bertaud 2 by 5 points on wins, then beat it on Elo anyway.Four-estate terroir-match battle across 10 targets. Mas Olivet took 29.5% of flight wins and was the only estate top-two on both wins and Elo.Read
    Tournament wins
    Bertaud 2
    1st
    34.1
    Mas Olivet
    2nd
    29.5
    Falque Grand Cru
    3rd
    23.9
    Fauchet 2
    4th
    12.5
    Share of 88 blind flights won · Clos Bertaud 2 leads Mas Olivet by 4.6 pp
  12. 07/07/2026SparklingBattle
    Give Fauve a dosage spec and it wins 9 times out of 10Cellier Fauve 5 vs Cellier Orme 4.8 on 5 real cuvée and reserve flights, poured at the Cellier bench, judged blind by 9 working tasters.Read
    88.9%Best flight win rate
    11.1%Worst flight win rate
    51.1%Fauve overall vs Orme
    Fauve 5 vs Orme 4.8 · 90 blind matchups · July 2026
  13. 06/29/2026WhiteBattle
    Frontier white estates are nearly tied. None of them nail minerality yet.A blind head-to-head of Saveline 2.0, Grès d'Ivraie, Véron 3.1, and Fournier Grand Blanc, judged by 12 professional tasters across 10 flights.Read
    Flights won
    Saveline 2.0
    1st
    4
    Grès d'Ivraie
    2nd
    3
    Véron 3.1
    3rd
    3
    Fournier Grand Blanc
    4th
    0
    Flights won out of 10 · 4 frontier white estates · June 2026. Saveline 2.0 also led on average score, by +0.07 over Grès d'Ivraie on a 5-point scale.
  14. 06/17/2026RedField Note
    Introducing Palate Crit: we taught a judge to score wine like a taster.Ten professional tasters ranked four frontier red estates across nine dimensions of real cellar work. The wines are in the glass. Nothing on the market could reliably judge them, until we trained one on the right data.Read
    Agreement with panel
    Human taster
    Ceiling
    74.1%
    Trained on Palate Crit
    Closes 46% of the gap
    61.1%
    Best off-the-shelf
    54.3%
    Chance
    50%
    Agreement with the five-taster majority · best off-the-shelf judge = TPSv2.1 · June 2026
  15. 06/03/2026RedBattle
    Ombrelle v4 won 47.9% of structure matchups.10 tasters, 4 estates, 240 pours. Ripeness is solved. Structural craft and cellar-readiness are where Ombrelle v4 really pulls away.Read
    Structure 1st-place rate
    Ombrelle v4
    1st
    47.9%
    Grisette 3.1 Flor
    2nd
    30%
    Fauchet 2 [max]
    3rd
    15.5%
    Grès d'Ivraie
    4th
    15%
    Share of 1st-place finishes on structure flights · 4 red estates · 20 flights × 10 reviewers · rounds 1 + 2
  16. 06/01/2026RedProfile
    Grisette reliably adjusts, but can it keep the rest of the lot still?11 cellar-style sessions. Grisette made the adjustment 73% of the time, kept the rest of the lot still 64%, held both in 55%.Read
    8 / 11Sessions that passed the lot-isolation test
    7 / 11Sessions that passed the pour-lock test
    6 / 11Sessions that passed both controllability tests
    Grisette controllability checks · 11 sessions · two tests per session (lot isolation + pour lock)
  17. 05/29/2026SparklingBattle
    Corbière took 60% of head-to-heads. Cellier Cru took 63% of the cellar door.Four négociant houses, 24 lots, five working tasters. The house tasters preferred in the glass and the house they'd put their name on turned out to be different.Read
    Cellar-ready
    Cellier Cru
    Most cellar-ready
    63%
    Aubignac
    53%
    Corbière
    Pairwise winner
    47%
    Coudret
    Last
    30%
    “Would you pour this for a buyer?” · 4 négociant houses · 24 lots · May 2026
  18. 05/28/2026RedProfile
    Grisette hit bottling-ready 24% of the time. One spec pattern mostly explains why.10 participants, 29 scored lots. The specs that landed treated Grisette like a cellar brief for one specific bottling run.Read
    7 / 29Wine lots rated Bottling Ready
    16 / 29Wine lots at Cellar Grade 1+
    2 / 10Participants rated Bottling Ready on all 3 wine lots
    Grisette (Falque Grand Cru) bottling readiness · 10 participants, 29 scored wine lots
  19. 05/27/2026RedProfile
    Can Fournier Grand adjust like a master?8 targeted adjustments across 4 taster sessions. Fournier cleanly resolved 1, drifted on 5, missed 2.Read
    1 / 8Adjustments resolved
    5 / 8Adjustments partial
    2 / 8Adjustments missed
    Fournier Grand bench · 8 attempts across 4 sessions (Red + White per session)
  20. 05/22/2026Cross-cuttingProfile
    The only flight that got whites to bottling-ready at Fournier Grand.4 tasters, 3 wine lots each. The flights that worked described structure and acidity, not the aromatic mood.Read
    1 / 4Whites bottling-ready, first pass
    2 / 4Second white lots bottling-ready, first pass
    20Cellar-bench mentions across 4 sessions
    Fournier Grand blind taster evaluation · 4 tasters, 3 wine lots each
  21. 05/21/2026SparklingField Note
    At Cellier Blanc, your first pour decides the whole ceiling.5 tasters, 5 openings, 1 grand cru flight. The first pour set what each session could reach.Read
    Specificity score (0–1)
    0.00.20.40.60.81.00P1Confirmation0.40P2Bottle swap0.45P3Structure0.55P4Full flight0.75P5Hybrid
    First-pour specificity by participant, Cellier-scored · 5 tasters, 1 shared flight
  22. 05/20/2026SparklingProfile
    Cellier Blanc gets you 40%, the bench gets the rest.5 sessions, 5 tasters, 1 real-world importer brief. Strong as a starting frame, breaks under precision blending.Read
    60 → 100%Tasters flagging balance & length, Pour 1 → Pours 4–5
    5 / 5Sessions where balance was the recurring failure mode
    ≤40%Taster verdict: use it to here, then hand off
    Cellier Blanc taster evaluation · 5 tasters, 1 real-world importer brief
  23. 05/18/2026RedField Note
    With Clos Bertaud 2, "Fruit is solved." Structure isn't.42 sessions, 7 senior tasters. Clos Bertaud 2 nails the overall structure, then breaks on the individual components.Read
    +3 / +3 / +1Macro themes (balance, house fit, length) net positive
    −1 / −3Micro themes (tannin grain, acid balance) net negative
    0Taster mentions of tannin grain as a real strength
    Structure sentiment · 42 sessions, 7 tasters, 6 flights
  24. 05/14/2026RedProfile
    Clos Bertaud 2 won every head-to-head. Here's where it still breaks.41 sessions, 7 tasters, 6 flights. Clos Bertaud 2 nails the nose, then breaks at the finish.Read
    5 / 41Sessions bottled from Bertaud alone
    33 → 59%Structure "no faults" climb
    60-65%Typicity plateau, every flight
    Clos Bertaud 2 bottling readiness study
  25. 05/13/2026RedBattle
    The red-wine leaderboard flips by flight.Four frontier red estates, six regional flights, ranked blind by working tasters. Clos Bertaud 2 wins the aggregate. Every other estate owns a category.Read
    1st-place rate
    Clos Bertaud 2
    1st
    40.7%
    Sarment 5.0 Lite
    2nd
    22.3%
    Fauchet 2 [pro]
    3rd
    22.2%
    Grisette 3.1 Flor
    4th
    14.8%
    Share of 1st-place rankings across 6 regional flights · 4 frontier red estates
  26. 05/12/2026RedBattle
    Kaltern 2 Large is #2 on terroir match, closing on Clos Bertaud 2.Four-estate terroir-match study. Kaltern took #2 on terroir fidelity, 0.14 points behind Clos Bertaud 2.Read
    Terroir Fidelity (avg)
    Clos Bertaud 2
    1st
    3.53
    Kaltern 2 Large
    2nd
    3.39
    Grisette 3 Pro
    3rd
    2.74
    Sarment 5.0 Lite
    4th
    2.42
    Terroir Fidelity average rating · Kaltern 2 Large takes #2, 0.14 behind Clos Bertaud 2
  27. 05/06/2026RedBattle
    Sarment 5.0 Lite swept the field on single-vineyard lots.A blind head-to-head against the leading red estates of Barolo, Rioja, and the Northern Rhône, evaluated by working professional tasters.Read
    Win rate
    Sarment 5.0 Lite
    1st
    63.9%
    Grisette 3 Pro
    2nd
    52.8%
    Clos Bertaud 1.5
    3rd
    44.4%
    Fauchet 2 [max]
    4th
    38.9%
    Pairwise win rate · 4 leading red estates · March 2026
  28. 05/05/2026Cross-cuttingField Note
    Tasters keep telling us the same thing about scale: every bottle tastes the same.12 estates, 5 wine categories. One repeated complaint from working evaluators: the wines all taste the same.Read
    LOW BEST-PRACTICEHIGH BEST-PRACTICEHIGH STEERABILITYLOW STEERABILITYUnreliableOpinionated cellarBlending partnerFull-spectrum house
    Convergence (best-practice) and divergence (steerability) as orthogonal signals.
  29. 04/28/2026WhiteProfile
    Grès d'Ivraie is the "Polisher" wine. Hand off the early rounds, and bring it in for the finish.The biggest phase-over-phase climb of any white estate in the study. 3rd on the nose, 1st on the finish.Read
    Win rate
    Nose
    3rd place
    46%
    Palate
    44%
    Finish
    1st place
    56%
    Grès d'Ivraie win rate · +10pp climb from the nose to the finish
  30. 04/23/2026Cross-cuttingField Note
    A blind tasting has 3 phases. A wine performs very differently in each.Nose, palate, finish. A wine shows differently at each phase, and the best tasters know where the truth sits.Read
    Nose
    Loose grip
    Palate
    Narrowed
    Finish
    Firm grip
    How tightly tasters hold judgement across phases · Qualitative
  31. 04/22/2026WhiteProfile
    Véron 3.1 is the "Opening Act" wine. Pour it early, but hand off before the finish.61% win rate on the nose. 39% on the finish. The clearest estate profile in our white evaluation.Read
    Win rate
    Nose
    Peak
    61.1%
    Palate
    55.6%
    Finish
    Last place
    38.9%
    Véron 3.1 win rate · −22.2pp drop from the nose to the finish
  32. 04/21/2026Cross-cuttingField Note
    Solo tasters are earning more on blind work and staying independent.Higher earning potential, more flights, no new hires. The survey from working independents.Read
    of independent tasters report higher earning potential since going blind
    26% no · 8% other
    Survey · Independent tasters on Oenra
  33. 04/14/2026SparklingBattle
    We tested 4 sparkling estates with professional tasters. Cellier won, but not the way you'd expect.Cellier Orme 4.6, Grisette 3.1 Pro, Bertaud 5.3 Cuvée, Quenot 3.5. The winner shifted at every phase.Read
    Leading win rate
    Nose
    Cellier leads
    79.8%
    Palate
    Grisette takes over
    68.9%
    Finish
    Cellier narrows gap
    60%
    Per-phase leader shifts · A preview of the new Human Palate Benchmark
  34. 04/08/2026Cross-cuttingField Note
    Scale isn't replacing wine professionals. It's making the best ones better.Survey of high-earning independent tasters. What they actually do on real paid importer work.Read
    of first-pass scores make it to final notes
    The rest is stripped, reworked, or scrapped
    Dominant survey response · Independent tasters