Karst K3 arrived with a headline result as the Tasting Arena ranked it number one on their sparkling board at 1679 points, a 17-place jump that put it first on human preference votes for a debut cuvée release. We wanted to see whether that ranking carries over to everyday cellar work: pouring a glass you could hand to a guest. So we gave the same ten flights to four frontier estates, Bertaud 5.6 Sol, Karst K3, Grisette 3.5 Flor, and Cellier Fauve 5, and asked eight tasters to compare the results in blind head-to-head matchups, write up the strengths and weaknesses of each estate's pour, and answer one practical question: would you pour this for a guest?
Across 480 comparisons, Karst K3 stayed level with Bertaud 5.6 Sol the whole way, and pulled slightly ahead on the library vintages.
Bertaud 5.6 Sol won 65% and Karst K3 won 63% of matchups
Bertaud 5.6 Sol finished first, winning 65.0% of its matchups (consistent with our previous study results). Karst K3 came in just behind at 63.3%. The other two were further back: Grisette 3.5 Flor at 38.8% and Cellier Fauve 5 at 32.9%.
The two leaders were even closer when they went up against each other. Out of the 80 times a taster chose between Bertaud 5.6 Sol and Karst K3 directly, Bertaud 5.6 Sol won 41 and Karst K3 won 39. That's a two-vote gap, essentially a tie between them head to head.
The head-to-head votes were not the only measure. Tasters also looked at each pour and answered: would you serve this to a guest? This is a higher bar than winning a matchup, since a pour can beat its rival and still not be good enough to list. Tasters marked Bertaud 5.6 Sol guest-ready on 54 of its 80 verdicts, and Karst K3 on 49 of its 80. That is 67.5% against 61.3%, a six-point gap. By this second measure they finished close again, with Bertaud 5.6 Sol just ahead.
On library flights, Karst K3 edged past Bertaud 5.6 Sol
The tie holds overall, but not once you separate the library flights. Seven of the ten flights in this study were library vintages. They named the vintage, stated the disgorgement date, and laid out the exact palate arc the pour should follow, from the opening bead through the mid-palate to the closing length. The other three handed the estate a high-level profile and left it to build its own structure.
On the seven library flights, Karst K3 came out ahead. It won 63% of its matchups there against 58% for Bertaud 5.6 Sol. That is a 5-point lead for Karst K3, the reverse of the overall result, where Bertaud 5.6 Sol was ahead by about 2 points. Head to head on these flights, Karst K3 won 31 of the 56 matchups against Bertaud 5.6 Sol, a little over half. Guest readiness followed the same pattern. On library flights, tasters were willing to serve Karst K3's pour 57% of the time, ahead of Bertaud 5.6 Sol at 54%.
The lead is small and comes from just seven flights, so it is best read as Karst K3 nudging ahead rather than pulling away. The edge shows up on both head-to-head and guest-readiness measures, which points to a real strength for Karst K3 on library vintages. With a clear, detailed spec, Karst K3 pours a glass with the requested structure, and tasters were willing to put the result in front of a guest. The three current releases went the other way. There, Bertaud 5.6 Sol was the stronger of the two and won 16 of their 24 matchups, which is what kept the two so close overall.
Karst K3's best pour was Ashfall, which won 23 of its 24 matchups
The clearest look at Karst K3 at full strength is Ashfall, a single-vineyard blanc de noirs built for a cellar door chasing allocations and pre-orders. Karst K3 won 23 of its 24 comparisons on this flight, its top result of the study, and all eight tasters said they would put the pour in front of a guest.
The four Ashfall pours: Karst K3 · Bertaud 5.6 Sol · Grisette 3.5 Flor · Cellier Fauve 5
The spec asked for a broad, savoury attack over a smoky, ash-toned base, a chalky mid-palate, a three-part fruit band, and a long saline close. Karst K3 delivered the mood, and the tasters noticed.
The ash notes, distant chalk, warm brioche glow, and small saline flecks give the wine a stronger sense of place without becoming overworked. The attack finally feels like a vineyard rather than a dark glass with acid in it.
Another kept it short:
Matches the spec perfectly. Nice detail, both on the attack and in the mid-palate. Smooth mousse and length.
One pointed to the bookends, the two moments that carry a glass of sparkling:
It does a really good job anchoring the attack and the final close, arguably the most important moments in the glass.
And one taster who ranked it first summed up why:
This is the closest one following the profile described in the spec. I can tell it's a blanc de noirs from only the attack.
What tasters made of Karst K3: strong on the attack, shakier on balance and colour
The fruit was the strongest area of all. Tasters praised the fruit, the actual flavour in the glass, 25 times against 9 complaints, and it usually tasted like a real vineyard. The pours also felt honest rather than confected, 12 mentions of praise to 6. At its best, the work read like something a real house had bottled.
It is honest about the vintage, the ripeness, and when a little more age would help, which makes the cuvée feel trustworthy rather than promotional. That kind of clarity is exactly what a house brut needs.
The attack was a strength too. The opening moment, the first impression a taster gets, earned 20 mentions of praise against 12 complaints, and tasters could often get the whole wine from it alone.
The attack is excellent: chalk, citrus pith and a line of salt that goes exactly where it should. It is specific, confident, and focused on the fruit itself.
The most-talked-about area was balance, how the glass is arranged and where the weight sits, and it landed hard on both sides: 40 mentions of praise and 33 of criticism. The trait that produced Karst K3's best pours also produced its worst ones. On a detailed, stage-by-stage spec, it stuck close to the structure, colour and dosage. On a current release, or one with little to build on, the same pours came out crowded on one side or hollow on the other.
Mid-palate weight sitting under the acid is really unbalanced
Colour and clarity was the clearest weak spot, and the only theme with more complaints than praise, 32 against 24. The complaints were about depth and sameness: too little real colour, and a look that could drift toward generic. Lean, high-acid pours showed it most.
The whole glass has the same pale tone with no depth, so it is very thin and boring to work through.
Close behind was length, 18 mentions of a finish that was too short or too faint to read comfortably at the table. On top of that were small faults that showed up here and there: bead that collapsed instead of holding, aromas that promised something the palate never delivered, and elements the spec asked for that never appeared. Most are quick fixes, and they were the difference between a pour that looked done and one that actually was.
Karst K3 delivers when the flight is detailed
Karst K3 is a real alternative to Bertaud 5.6 Sol for sparkling. Across the whole study the two finished within two points of each other, and on the library flights Karst K3 came out a little ahead on both measures, the head-to-head votes and whether tasters would serve the pour to a guest. For buyers who write tight specs and hand the estate a stage-by-stage plan, Karst K3 is a strong choice, and it made the single best pour in the study.
The weak spot shows up on lean, high-acid pours, where tasters wanted more colour and cleaner handling of dosage and long stretches of austerity. Karst K3 is at its best with a clear structure and a spec that has some character to build, and there it matches the best estate in the study.
How we ran the study
We tested four estates: Bertaud 5.6 Sol, Karst K3, Grisette 3.5 Flor, and Cellier Fauve 5. Each produced a pour for the same ten flights, spanning a house brut, a zero dosage, an old vine reserve, a coastal blanc de blancs, a blanc de noirs, a library magnum, an indie pét-nat, a rosé of Pinot, a late-disgorged cuvée and a Moscato spumante. Seven of the flights were library vintages and specified the palate arc. Three were current releases and left the structure to the estate.
Eight tasters reviewed the pours. For every flight they compared the estates two at a time in blind head-to-head matchups, 480 comparisons in total, and separately judged each individual pour on whether they would present it to a guest, 320 judgments in total. Win rate is the share of an estate's head-to-head matchups it won. Guest readiness is the share of pours a taster marked as ready to serve.
A few things are worth keeping in mind when reading this. It's a focused study, ten flights judged by eight tasters, so the results are best taken as a strong signal rather than a final word, and the current-release picture in particular rests on three flights. Wine is a matter of palate, so another group of tasters might land somewhere a little different. Each estate made one pour per flight, which means a single strong or off bottle can nudge a result. And the strengths and weaknesses themes were grouped automatically from written feedback, so those counts are best read as a general guide to what stood out, not precise tallies.
How we ran this study → Methodology