2026 · Round 21 · Las Vegas Grand Prix

Las Vegas briefing, Saturday edition: Antonelli fastest but sixth on the grid, Verstappen on pole, and the battery runs dry 1 km before the end of the Strip

Briefing

Antonelli set the fastest lap in qualifying, 1:33.893, and starts sixth after a five-place engine penalty. Verstappen inherits pole from Russell and Norris. Piastri qualified seventh and starts seventeenth. On the Strip the cars peaked at 336 km/h and then lost 39 km/h on full throttle before the Turn 14 braking point, more than any simulation had predicted. The race is forecast at 7 °C, five degrees colder than on Thursday. Nobody has run the hard (C3) on a dry track all weekend, and FP3 showed no graining on a track that was still getting faster.

Our planOne stop. Soft (C5) → medium (C4) around lap 26 (window 18–31) or medium (C4) → soft (C5) around lap 24, 0.7 s apart over the race; the best two-stop is 21 s slower. Pit loss 22.5 s. The hard (C3) is untested in the dry and stays out of our plan.
Biggest riskThe cold. 7 °C in the air is colder than anything the 2026 tyre has raced in, and FP3 could not show graining because the track gained 2–4 s per run as it rubbered in. A tyre that grains for real moves the race to two stops. Second risk: the front-row tyre count, with Verstappen, Hadjar, Lindblad and Lawson on one new soft (C5) each.
Watch forThe first ten laps of the first stint: if lap times climb by a second or more within a few laps, it is 2024 again. Then Turn 16 every lap: a car within 1.0 s there has Overtake down the Strip a lap later.
Strategy

Qualifying, at 15.8 °C track, went to Antonelli by 0.258 s: 1:33.893 against Verstappen's 1:34.151, with Russell +0.331, Norris +0.351, Hadjar +0.699, Lindblad +0.775, Piastri +1.006, Lawson +1.197, Gasly +1.212 and Colapinto +1.421. Leclerc and Hamilton went out in Q2, eleventh and twelfth. Then the penalties: Antonelli drops five for a sixth engine, Piastri and Sainz ten each, Stroll ten, Bearman to the back. The FIA's provisional grid: Verstappen, Russell, Norris, Hadjar, Lindblad, Antonelli, Lawson, Gasly, Colapinto, Leclerc, Hamilton, Hülkenberg, Bortoleto, Albon, Câmara, Pérez, Piastri, Alonso, Bottas, Sainz, Stroll, Bearman. The stewards gave Verstappen a warning for his driving at Turn 14 and took no further action over Turn 17; an incident between Verstappen and Russell is still under investigation.

Our planner, on FP3's long runs and the season's wear for this kind of surface: one stop, soft (C5) → medium (C4) with the stop around lap 26 and a 1-second window from lap 18 to 31, or medium (C4) → soft (C5) around lap 24, 0.7 s slower. The best two-stop is 20.7 s behind. The hard (C3) is not in the plan because nobody has run it on a dry track this weekend. Our field simulation on this grid is not worth publishing: FP3's driver levels are mixed up by a track that gained 2–4 s per run, and it put Verstappen 15th from pole. We read the front from qualifying instead, where Antonelli was 0.26 s clear.

That leaves the tyre count as the strategic story. On our count of new sets, Verstappen, Hadjar, Lindblad and Lawson have one new soft (C5) left each, Gasly none, Russell, Norris, Antonelli and Piastri two, Leclerc and Hamilton four. Everyone has two new hards (C3) and two or three new mediums (C4). The cars at the front start on used softs (C5) or new mediums (C4), and the two Ferraris from tenth and eleventh are the only front-half cars that can run a new soft (C5) in every stint.

The championship was settled in Mexico, so Las Vegas is about the last three race wins of the season and a constructors' table where Mercedes leads Ferrari by 142 points with 129 left. After São Paulo: Antonelli 390, Verstappen 279, Russell 262, Hamilton 261, Leclerc 249, Norris 246.

Pirelli predicts one stop and says the soft (C5) can be a race tyre. Both statements fit what this car has done all autumn on the same three compounds. In Mexico Antonelli ran 43 laps on one soft (C5) with no measurable wear. In São Paulo, on a 25 °C track, the soft (C5) ran 734 clean laps with lap times that fell with age, Russell and Leclerc did 45 laps on one set, and no car fitted a medium (C4) or a hard (C3). We planned São Paulo around the medium (C4) because Friday said so, and we were wrong; the soft (C5) was the race tyre.

Las Vegas is the test of whether that holds in real cold. The forecast for FP2 and qualifying is 7–10 °C in the air, colder than anything the 2026 tyre has raced in, and the one weekend this year that came close, Shanghai in March, is the one Racing Bulls' Guillaume Ducreux names as the only case of severe graining so far. Graining is not wear. It is rubber tearing off the surface and sticking back on, which costs grip for a phase of the stint and then, usually, clears. It turns the plan from a stint-length question into a timing question: when in the stint the grip goes, and whether a driver waits for it to come back or stops. Ducreux puts it simply: 2024 was two stops because of graining, 2025 one stop because there was little.

So the plan before FP1 is the plan without numbers. One stop. The second compound is the one that switches on fastest in the cold, and on this year's evidence that is the medium (C4) rather than the hard (C3). We will not set stop windows before Friday's long runs, because this year's Friday wear has overstated race wear by between three and seven times at every circuit since Austin, and by more than that in São Paulo, and a cold Friday night is the session most likely to overstate it again.

The pit stop is the strategic oddity. Pirelli says the circuit reduces the effect of the traditional undercut, because a new set cools on the long run from the pit exit before it reaches the corners where it would gain. A car on older, warm tyres can be quicker for a lap than the car that has just stopped. If that is right, the overcut is the stronger move here, and the first car to stop in a close fight is the one that loses. The pit loss itself is a track property and our measurements stand: 22.1 s in 2024, 22.9 s in 2025. Under a VSC in 2025 it was about 13.5 s, as Antonelli's lap-2 stop showed.

Tyres

Hard (C3), medium (C4), soft (C5), the softest three in the range, with the hard (C3) and the medium (C4) as the mandatory race tyres and the soft (C5) as the Q3 tyre. Allocation 2 hard (C3), 3 medium (C4), 8 soft (C5), plus five sets of intermediates and two of wets. Pirelli's prescriptions are unusual in one way: minimum starting pressure 27.5 psi on the front and 28.0 psi on the rear, the rear higher than the front, where São Paulo had 27.0 and 26.5. Camber −3.0° front and −1.75° rear. Fitted tyres may not be left in air colder than 10 °C, which in Las Vegas at night is a rule teams will have to work around in the garage.

Pirelli's own notes: graining likely because of low grip and cold, degradation mainly mechanical, tyre loads high only in the middle sector, lock-ups and flat spots a risk, warm-up a key factor. The front-right is the tyre that has grained most here, according to F1.com's preview, and some teams run an asymmetric set-up to protect it.

What we measure is the shape of the first laps of a stint as well as the slope. A tyre that grains shows lap times that rise over a few laps and then, if the stint is long enough, fall again as the surface cleans up; a tyre that wears shows a steady rise. Marek Lubieniecki's chart of the first stint on the medium (C4) here, the same compound in both years, shows the shape: in 2024 the top ten were 1.7 s slower by lap 9 than on laps 5 to 7 and stopped soon after, in 2025 they got faster lap after lap as the fuel burned off. Our own race data from those years says the same, and we use it only for the shape, not for this year's numbers. FP2 was run on wet and intermediate tyres, so the first dry long runs come in FP3, a single hour that also has to hold the qualifying runs. Runs of six to eight laps will be enough to see a climb of a second or more, which is the 2024 signature, but not long enough to see the tyre come back. Our wear fit reads the slope across the stint, which mixes the two, so FP3's runs get read one by one before anything goes into the planner. The library holds Las Vegas wear from 2024 and 2025, and we do not use it: the 2026 tyres are a new construction and last year's stint lengths say nothing about this year's.

FP3 was the first dry hour of the weekend, at 17 °C track, and it could not answer the graining question. Lap times fell by 2 to 4 s over runs of ten laps as the track rubbered in, which hides any graining underneath. Russell's medium (C4) run went 97.30 on lap 9 and then 97.46, 97.75, 97.89, a rise of 0.6 s over three laps that is the only hint in the session. Nobody showed the 2024 signature of a second or more within a few laps. The soft (C5) and the medium (C4) ran at the same level on long runs, 99.10 and 99.16 s. Nobody ran the hard (C3).

Energy and power unit

The FIA's power unit document for Las Vegas: 8.0 MJ of recharge per lap in the race, 8.5 with overtake active, 8.5 in practice, 9.0 on out-laps, and 6.0 MJ in qualifying – the lowest qualifying figure of the autumn, against 6.5 in São Paulo. The alternative curve, a 250 kW cap up to about 310 km/h, applies in the race at T1–T3 and T5–T12, 1,830 m or 29.5 percent of the lap. The Strip from T12 to T14 and the start/finish straight from T16 to T1 run on the full 350 kW base curve. The power curves are the same as in Mexico and São Paulo: full 350 kW to 290 km/h, falling to 100 kW at 340 and to zero at 345.

That last number matters here more than anywhere. The cars reached around 340 km/h on the Strip in previous years with DRS, and on this year's rules the MGU-K stops at 345 km/h on its own. Some of the speed plateau we will see at the end of the Strip is the rule, not the battery. The rest, in qualifying with 6.0 MJ, is superclipping, and Las Vegas should show it more clearly than any track this year. We will measure the speed fade over the last 500 m of the Strip in FP1 and FP2, as at every circuit since Monza, and separate lifting from clipping with the throttle trace.

Overtake is decided at one place on the lap: the detection line at 5,299 m, before Turn 16. A car less than 1.0 s behind there gets Overtake from the activation line at 5,400 m, and under the sporting regulations (B7.2.3) it stays active for the whole lap, until the car next crosses the activation line more than a second behind. So the chaser carries it through T17, down the start/finish straight to T1 and up the 1.9 km Strip to T14. What it buys is power at top speed. On the standard curve the MGU-K gives 350 kW to 290 km/h and then 5 kW less for every km/h, 175 kW at 325 km/h; with Overtake it holds 350 kW to 338 km/h. At the end of the Strip the car behind has twice the electric power of the car in front, and 0.5 MJ more recharge for the lap. Marek Lubieniecki's simulation puts it at 338 km/h against 325 in clean air, and about 8 km/h more for a real chaser once the dirty air out of Turn 12 is counted. His calibration on this year's Monza and Spa: from one second back the chaser is still about 60 m behind at the braking point and passes once in about 50 tries; from half a second it is about 13 m back and passes about one time in three. In Mexico, on the same rules, 8.8 percent of close fights ended in a pass; here it was 19.8 percent in 2024 and 13.8 in 2025.

Qualifying gave us the measurement. On the Strip, between 3,270 and 5,020 m, the cars reached a median 336 km/h about 1,000 m before the Turn 14 braking point and then lost a median 39 km/h on full throttle, to 297 km/h. Marek Lubieniecki had simulated 341 km/h falling to 316 with the power gone 790 m before braking; the real fade started earlier and went further. The spread is the story: Norris lost 14.5 km/h and Piastri 15 from a lower peak of 331–332 km/h, while Russell lost 49.5, Verstappen 47, Hadjar 47 and Bottas 56. McLaren spent its energy differently, reaching the end of the Strip 25 km/h faster than Russell and Verstappen, and Norris qualified 0.02 s behind Russell. On the start/finish straight the fade was 19 km/h. In the race the rules give 8.0 MJ, not 6.0, and a lower power curve on the straights, so the race fade will be smaller and mostly the rule rather than the battery.

Aerodynamics and car

The thinnest rear wings of the season come out here, and the FIA's car presentation submissions show how far Mercedes and Ferrari have gone. Mercedes declares four changes, all aimed at the straights: the winglets removed from the rear wing and the mirror stays trimmed, both brought back from Monza, the winglets at the rear corners repositioned and shortened, and a narrower exit at the back of the engine cover, which Mercedes says is enough cooling for Las Vegas. Ferrari declares one: the front floor board with a single vertical element, the same part it raced at Monza. The other three low-drag items Ferrari brought to Monza – shorter mirror stays, the brake-duct winglets removed, the slotted tail wing – are not on the list, and Ferrari does not declare a cooling change, having opened up the engine cover for the altitude in Mexico. Mercedes has trimmed the car for Las Vegas; Ferrari has fitted one part of its Monza kit. The rest of the field splits into three groups. Racing Bulls (three items), Audi and Haas (one each) also trim for the straights: Racing Bulls with shorter floor-board elements, different winglets on the rear brake drums and a low-load front-wing flap to keep the balance once the rear has lost load; Audi with a new rear brake-duct deflector; Haas with a revised rear wing that changes how its Straight Mode opens, declared as an evaluation. McLaren goes the other way, with the only parts of the weekend built for downforce rather than for this circuit: a new floor board, floor edge and rear floor furniture for flow and load. Red Bull lowers and shortens the dive-plane on its front-wing endplate. Alpine, Williams, Aston Martin and Cadillac declare nothing new, which for Alpine means the car that has topped the qualifying speed trap six times this season arrives without a single Las Vegas part. F1.com's preview notes that the cold gives free power-unit cooling, so bodywork can be tight, and that the brakes need to keep heat rather than shed it, with some teams sealing the brake drums so the carbon is warm enough at the end of the Strip. Brake heat is also what warms the tyre carcass, which makes brake cooling part of the tyre problem here. Mark Hughes and Giorgio Piola report that Brembo has tested a new disc on the Red Bull in Austin and Mexico, possibly with blind holes on the inner face to keep the disc temperature more even. If it is photographed this weekend, a cold track with a 1.9 km straight is exactly where it would earn its place.

Meteorology

Our forecast from open-meteo for the session hours, local time: FP1 on Thursday at 16:30, 12–13 °C with a 30 percent chance of a shower and wind 15–21 km/h; FP2 at 20:00, 9–10 °C, dry, wind 25–34 km/h, the windiest session of the weekend. Qualifying on Friday at 20:00, 7–8 °C under a clear sky, the coldest. The race on Saturday at 20:00, 12–13 °C under cloud, dry, almost no wind. The race is forecast five degrees warmer than qualifying, and FP2 sits between them, which makes it the more useful guide to race tyre behaviour. Track temperature is not in the forecast; at night with no sun on the asphalt it will sit only a few degrees above the air.

Saturday's forecast has moved: 7–8 °C at the start at 20:00, overcast, dry, wind under 10 km/h. That is five degrees colder than Thursday's forecast for the race and the same as qualifying. Track temperature in qualifying was 15.8 °C.

Track and pit lane

6.201 km, 50 laps, 17 corners, anticlockwise, about 610 m above sea level. The Race Director's notes list the changes since 2025: full-width resurfacing and micromilling at the exit of T13 and between T16 and T17, asphalt patching before and at T13 and before T14, painted kerbs removed at the exits of T4 and T5, five manhole covers replaced by retrofit lids, and the pit-entry wall realigned. Resurfaced sections are smoother and, in the cold, lower in grip for the first sessions. On a resumption the Safety Car waits at T13, and there are no waved yellow flags or SC boards: light panels only.

Neutralisations, from our race data: two Safety Cars and a VSC in 2023, nothing in 2024, a short VSC early in 2025. That is less than the "high chance of a Safety Car" the street-circuit label suggests. Overtaking, by our encounter count: 45 passes from 227 times a car ran within reach of the one ahead in 2024, 19.8 percent, and 23 from 167 in 2025, 13.8 percent. For comparison, this season's median is 16.5 percent and Mexico had 8.8. Passing is possible here, but not cheap.

Rivals

Mercedes won here in 2024 with a one–two, Russell ahead of Hamilton, and finished second and third in 2025 behind Verstappen after both McLarens were disqualified for plank wear. Antonelli's 2025 race is the reference for an offset strategy: from 17th, the only car on the soft (C5), a stop on lap 2 under the VSC for 13.5 s instead of 20, and 48 laps on the hard (C3) to finish third. The plank disqualifications are the other reference: the Las Vegas surface is bumpy, and running the car low for the straights carries a cost that is checked after the flag.

Form on the same three compounds this autumn: in São Paulo McLaren had the fastest race pace, Mercedes 0.04 s per lap behind, Red Bull 0.26–0.40 and Ferrari 0.42–0.59.

Drivers

Norris, who won in São Paulo, on Las Vegas: one of the hardest races of the year, little room for error, white lines that are bad in the cold, graining that is hard to understand, and a track that does not dry quickly when it rains. Verstappen comes from the 21st-to-second drive in São Paulo and has won here twice. Antonelli is champion and knows the offset strategy from last year.

Rules and stewards

Under the 2026 rules a dry race needs two slick compounds; São Paulo was exempt because the intermediate was used. The timetable in Swedish time: FP1 Friday 01.30, FP2 Friday 05.00, FP3 Saturday 01.30, qualifying Saturday 05.00, the race Sunday 05.00.

What would make us wrong

A Friday that grains and a Saturday that does not. FP2 is forecast three degrees colder than the race, so a stint that grains on Thursday night could run clean on Saturday, and a plan built on Friday's graining would stop once too often. The opposite case is a soft (C5) that grains for real in race conditions, which would push the race towards the medium (C4) and the hard (C3) and a second stop. And the undercut: if the out-lap loss is smaller than Pirelli expects on this year's car, the overcut advantage disappears and the order of the stops reverts to normal.

Saturday adds one. The cold race forecast makes the tyre count matter more than the planner says: a soft (C5) that does not switch on after a stop at 7 °C hands the overcut to the car that stayed out, which is what Pirelli warned about, and the cars with one new soft (C5) left have the least room to react.

How to read this: the briefing is written before the race from practice and qualifying data, FIA documents and Pirelli's information. The sections mirror how a Formula 1 team splits the work; the numbers are ours and each carries its source. After the race the Aftermath says where we were right and where we were wrong.