Mexico City briefing, Thursday edition: new asphalt, thin air, and the one tyre question that matters
Mexico City is the race where the air does a quarter of the work less. At 2,285 m the cars lose about a quarter of their downforce and drag, the turbos and the batteries carry the engine, and the tyres slide more than the corners deserve. This year 68 % of the lap has new asphalt, Pirelli brings its three softest compounds, and the first practice session will be driven by ten rookies. We go in with a one-stop as the working plan, a library pit loss of 22 s, and a Friday evening that has to answer one question: does the C5 grain on the new surface, or does it race.
Seventy-one laps of 4.304 km at 2,285 m above sea level, race start 14:00 local, 21:00 in Sweden. The library has Mexico City at 22.0 s pit loss under green from 26 stops in 2025 (23.7 in 2024), 3.1 s under a VSC, and an overtaking threshold of 0.82 s per lap – the same as Austin – from 493 close encounters over two races, with a pass rate of 19–23 %. Six neutralised laps in 2024, two in 2025. Nothing in those numbers is about tyres, and nothing about tyres from earlier seasons goes into this plan: 2026 is a new regulation and the C3/C4/C5 trio has its own record this year, below.
The working plan is one stop. Pirelli's preview says so in plain words – a low-severity circuit, strategy decided "more by the confidence a driver can extract from a particular compound than by any genuine tyre life limitation" – and the season's record backs them: Pirelli has called the number of stops right in 16 of 18 races this year, and in Austin their medium → hard window of 21–27 contained the winner's stop on lap 24 while we argued for something else. Our planner on the library's wear level for this type of circuit, 0.035 s per lap, puts a one-stop with the switch around lap 25–32 of 71 at the front of the list, with medium → hard and soft → medium within a few seconds of each other. That answer is a placeholder until FP2. What replaces it is the slope of the soft and medium long runs on Friday evening, read from lap six onwards, and the first lesson of the autumn applies: Fridays have overstated race wear by three to seven times this season, so a Friday slope is an upper bound, not a forecast.
The strategic disagreement worth having is about the C5. Pirelli's preview says the soft "has demonstrated strong characteristics on low-severity circuits and could prove an attractive option not only in qualifying but also in the race". If that holds, the race is soft → medium or soft → hard from the start, which is what suits the cars that launch well on the soft. If the new asphalt grains it, the race is medium → hard with the soft spent on Saturday, which is what Mercedes ran in Austin and won with. Friday decides which race we are planning.
The compounds are C3 hard, C4 medium and C5 soft, with one of C3 or C4 mandatory in the race and the C5 as the Q3 tyre. The allocation is the usual two hard, three medium and eight soft; one soft set is reserved for Q3 and the Q3 drivers return one soft set after qualifying, so a Q1 or Q2 elimination leaves a driver with one more new soft for Sunday than the men at the front. The minimum pressures are the lowest of the autumn, 25.5 psi front and 24.5 rear against Austin's 26.5/25.5: with a quarter of the aerodynamic load missing, Pirelli expects less heat in the carcass, not more.
What this trio has done in 2026, from our library, on the eight dry races it has run: Melbourne, soft 0.059 s per lap of age, medium 0.005, hard 0.040, with the hard going 44 laps; Miami 0.036 / 0.028 / 0.047; Montreal 0.005 across the board over 68 laps, the lowest wear of the year; Monaco 0.035 / 0.045 / 0.036; Austria 0.067 / 0.082 / 0.077 on a 55 °C track, the one race this year where the trio forced three stops; Hungary 0.041 / 0.046 / 0.063; Monza 0.030 / 0.018 / 0.009 with the hard good for 50 laps; Singapore 0.041 / 0.037 on soft and medium. The C5 has been a race tyre at Montreal, Monza and Singapore and a sprint tyre at Austria and Melbourne. Mexico's low lateral loads put it in the first group on paper; the new asphalt and the thin air are what could move it to the second.
Graining is the mechanism. With less downforce the tyre slides more for the same corner speed, the surface tears before it wears, and the rear axle takes the worst of it on the traction out of the slow corners in the stadium. A new, smooth surface makes that more likely early in the weekend and less likely by Sunday as rubber goes down; Pirelli rates the track evolution at five out of five, its maximum. The practical reading: a soft that grains in FP1 may not grain in the race, and a soft that does not grain in FP2 is a race tyre. We measure the slope of the second half of each long run, not the first, because graining shows as a rising slope that then flattens when the surface cleans up.
One number to carry from Austin: a used Q1 soft started Hamilton's race with no measurable penalty against Leclerc's new one, equal slopes and two tenths of level. On a track where the soft may be the start tyre again, that matters for the drivers who spend their softs on Saturday.
At 2,285 m the air is about 23 % less dense than at sea level. Three things follow for a 2026 car. The turbocharger has to spin far faster to deliver the same boost, which is why Mexico has always been the hardest race on turbos and why teams run their most conservative engine maps on Friday. Cooling loses the same quarter of its capacity, so the bodywork is opened up and the cars run with more drag than their wings suggest. And the MGU-K's 350 kW does not care about air density at all, so the electrical share of the lap is worth more here than anywhere else: the combustion engine is weaker, the battery is not.
That changes the back-straight picture we measured in Austin. The main straight here is 1.2 km and, as Eloy Aparicio points out in his preview, the cars reach 350 km/h with maximum-downforce wings on, because drag is a quarter lower. The FIA's power-unit document for the weekend, when it comes, will say where the 350 kW taper starts and whether there are alternative deployment zones; on the Austin curve, full power tapers from about 290 km/h to zero at 345, which here would be reached well before the braking point for Turn 1. We will measure the fade on the main straight in FP2 with the same method as Austin, and expect larger drops than Austin's 6–21 km/h simply because the straight is longer and the terminal speed higher. Energy harvesting is the other side: the slow stadium section and the braking into Turns 1, 4 and 12 are the lap's recharge zones, and the 8.5 MJ race recharge limit will be a real constraint on a lap with one long straight and many short ones. Williams have flagged turbo pressure and the electrical compensation of the torque gap as the central challenge for the 2026 power units here; we will quote them when we have the text.
Ferrari's Austin package – reshaped floor leading edge, new floor boards, more volume at the lower front of the sidepod, a larger and steeper diffuser – is, in Mark Hughes's and Giorgio Piola's reading on F1.com, about keeping the airflow attached under the diffuser at low speed and high ride height, which is exactly the regime that Mexico's thin air stresses hardest: a quarter less flow energy through the diffuser, and the stadium section is the slow, high-ride-height case. Austin gave Ferrari the fastest race car on our form measure for the first time since Barcelona. If the new floor holds its downforce here, that is a stronger proof than Austin was. Mercedes brought nothing new to Austin and won; their season total of 38 new parts is the fewest of the top four (Red Bull 56, McLaren 54, Ferrari 50). Whatever arrives on Thursday goes into the FIA car presentation list and onto our upgrades page on Friday.
The form picture after Austin, on fuel-corrected clean-lap race pace: Mercedes still the fastest car over the season on a weighted average, 0.22 s per lap from the front, but Ferrari is 0.04 behind them, Red Bull 0.09, McLaren 0.17, and in the last three races Mercedes has not been the fastest car once – Red Bull was quicker in Sepang, all three in Singapore, Ferrari and Red Bull in Austin. Eloy Aparicio's comparison of speed profiles puts Miami, Shanghai and Montreal closest to Mexico, and Mercedes took pole at all three; that was the spring car, and the form line says the gap has closed since. His provisional pole time is 1:17.386.
The five-day forecast is run on Friday morning and goes in here. What we look for is the afternoon track temperature in the sessions, and the afternoon shower risk that this altitude is known for. Track temperature in the afternoon sessions is what matters for graining: a cold, green track grains, a warm, rubbered one does not.
Pit loss 22.0 s under green in our library (2025, 26 stops), the shortest pit lane of the autumn after Austin's 21.2, and 3.1 s under a VSC. The new asphalt covers Turn 16 to Turn 1 and Turn 7 to Turn 16, leaving the first sector largely as before, so the braking into Turn 1 at the end of the 1.2 km straight is on the old surface while the exit and the whole stadium are new. Pirelli warns of lock-ups into Turn 1 early in the weekend, and with a quarter less downforce the braking zone there is the longest on the calendar. Overtaking threshold 0.82 s per lap of pace advantage, the same as Austin, from two races; the pass rate per encounter was 23 % in 2024 and 19 % in 2025. Two neutralised laps last year, six the year before. The undercut at 22 s is strong on a track where the out-lap on a soft is quick and the in-lap through the stadium is slow.
Ferrari is the only team that keeps both race drivers in FP1; Red Bull runs Ayumu Iwasa, McLaren Leonardo Fornaroli, Mercedes Frederik Vesti, Racing Bulls Nikola Tsolov, Audi Freddie Slater, Alpine Gabriele Minì, Cadillac Colton Herta, Williams Luke Browning, Aston Martin Jak Crawford and Haas Ryo Hirakawa. Ten rookies on a green, resurfaced track means FP1 is close to useless as a measure of the regular cars, and Ferrari gets a full Friday on the new surface that nobody else has. The head-to-heads from F1.com after Austin: Antonelli leads Russell 11–7 in qualifying and 13–5 in races; Leclerc and Hamilton are 10–8 and 9–9, the closest pair in the field; Norris leads Piastri 11–7 and 10–7; Verstappen leads Hadjar 11–4 and 12–3 despite Hadjar's third on the Austin grid. Pérez has his home race and leads Bottas 12–4 on Sundays at Cadillac.
The tyre count will decide the Sunday options again. In Austin, Red Bull's two new softs became Verstappen's two-stop and Mercedes' empty soft shelf became medium → hard; the same count on Saturday evening will say who can start on a new C5 here.
Driver form on race pace after Austin, weighted to recent races: Leclerc 0.24 s per lap from the fastest car, Verstappen 0.27, Antonelli 0.32, Hamilton 0.49, Norris 0.55, Piastri 0.70, Russell 0.73, Hadjar 0.86. Antonelli's start in Austin – 66 m in three seconds, the best launch in the field, on the medium – is the number to remember on a track where Turn 1 is 800 m from pole and the launch decides whether anyone arrives alongside.
After Austin: Antonelli 365, Russell 242, Hamilton 241, Verstappen 236, Leclerc 233, Norris 211. Six races left including this one; the title cannot be decided in Mexico. Constructors: Mercedes 607, Ferrari 474, McLaren and Red Bull level on 355. Pirelli's C5 is the Q3 tyre; Q1 and Q2 eliminations keep one more new soft for the race.
A C5 that races. If the soft holds a 25-lap stint on the new asphalt without graining, the fastest race is soft → medium and the teams with new softs left after qualifying have the plan to themselves. A C5 that grains in the race makes it medium → hard for everyone and turns Saturday's softs into a cost. And the Friday numbers themselves: on this season's record the race will wear the tyres at a fraction of what FP2 shows, and the mistake to avoid is the Austin one – believing the Friday number about the compound we have least data on.
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.