[Settings] My settings for hypercars endurance racing in single player

[Sparviero]

LMU is an excellent simulator for long-duration races (e.g., 3 hours or more), but it has some AI limitations that can compromise the fun, often tipping the race in favor of the player. Below are some settings that have improved my experience slightly while I wait for the game to develop further.

1. As others have reported, it's important to set the right AI strength level, using your qualifying time as a reference. The goal is to be about 0.5 seconds from pole (about 0.8-1.0 at Le Mans). Be careful because the level changes by about 2-3 points depending on the circuit. For me, 94 is generally a good value, but for Sebring and Bahrain I have to use 95, 97 at Fuji. It can also depend on the car (e.g., the 9x8 without wings requires about 2 more AI strength points).

2. Aggression must be at its maximum, otherwise the AI will waste too much time in lapping cars. However, be careful to balance this with the AI's strength. If the latter is too high, the high aggression will lead to collisions with you when overtaking. In this case, lower the AI, not the aggression. It's also recommended to only change the aggression to low at the start for one lap (fortunately, the system allows for changes during the game) to limit the risk of accidents at the start.

3. You need to change the fuel usage (AI Fuel Mult) in the settings.json file to make the AI's stints as long as yours, which is just over 45 minutes or an hour, depending on the circuit. 0.92 is a good starting point, but it depends on the circuit. Sebring 0.95, Portimao 0.83, Spa 0.93, Le Mans 0.92, Monza 0.92, Fuji 0.93, Bahrain 0.93, Qatar 0.83, Imola 0.93, San Paolo 0.92, COTA 0.92.

4. The AI doesn't manage tires well, over-wearing them and making unnecessary changes (e.g., putting on used tires). Therefore, it tends to make more pit stops with tire changes than the player. This destroys race strategy, as the tire change results in about 25-35 seconds of time lost between changing the tires and warming up for the next 2-3 laps, to the player's advantage. The best thing to do is activate the tire warming option. The result is that the AI is much more consistent in lap times, and the impact of the tire change is minimal (keep in mind that the AI is still a little quicker than you in the pit stop and is a little faster than normal in the first laps).

5. Set tires limit according to real life: 18 for 6H, 26 for 8H, 32 for 12H, 56 for 24H. If you ran half race you can use 9 for 3H, 13 for 4H, 16 for 5H, 28 for 12H. If you use more tires, AI will change them more frequently.

6. Shorten the first stint of Le Mans to 11 laps. This will put you out of sync with the AI without impacting your strategy, but will prevent the AI from DNFing when loading a saved game when the save happens with the AI in the pit lane.

 

[ndslpsp]

At Imola, during the start phase, with heavy fuel and cold tires, I was running 1:48. After a few laps, the tires warmed up and I started getting into the 1:47s. In the middle of the stint, as fuel gradually burned off, I dropped into the 1:46s, and a few laps before my pit stop, I even put in a 1:45. Okay, now let's look at the AI. With difficulty set to 95%, they came out of the gate doing 1:46s and quickly pulled away from me, while the cars behind me got stuck and couldn't get past.

By mid-stint, the AI started slowing down to 1:47s, and I gradually caught up and even overtook them. A few laps before their pit stops, they had already slowed to 1:48s. Within just a few laps, I had passed nearly every car. Coming out of the pits, they were busy changing tires while I was already starting to lap cars in my own class...

Well, you might say the difficulty is too low, so I crank it up further. But then, right after the start, they blast away from me, only for me to slowly chase them down again. After they pit and change tires, I overtake them once more. The process is never in sync. Yes, with the fuel multiplier dialed in, they run the same 27 laps per stint as I do—but so what? Within the same stint, I am always getting faster, while they are always getting slower. There's no way for us to be in sync.

The only thing I can do is raise the difficulty so that their raw pace over a full lap compensates for the time they lose from slowing down and from pit stops. But then their single-lap pace will inevitably be vastly superior to mine, and I'll just keep getting dropped on track with no real chance to fight them.

If you want a balanced race,Don't worry about whether there's a rival who can keep pace with you throughout the entire race—that's simply not achievable. the only real solution is to manipulate the total race distance...

 

[ndslpsp]

Through the adjustable fuel multiplier, single-player players are presented with only two types of endurance races:

1. Adjust the AI fuel multiplier so that the AI has lower fuel consumption, runs more laps per stint, and stays on track longer. At the same time, lower the difficulty so the AI is slower than you. Your race experience will be: constantly pitting, constantly falling behind, and constantly overtaking again, while doing your best to save fuel and tires to reduce pit stops and run more laps per stint.
2. Adjust the AI fuel multiplier to exactly match the player's consumption (the standard is that at the end of each stint, the AI near your position has almost the same remaining fuel as you). Increase the difficulty so the AI starts off fast, gains time on track early, then slows down at the end of each stint and wastes time with tire changes in the pits. Your race strategy will rely on gaining positions through pit-stop advantages, defending and holding off the AI in the early phase after their pit stops, and pushing hard to chase them down during the late phase of the stint when they slow down.

The calibration of both approaches still ultimately comes down to the hard standard of total race distance—Imola 200 laps for first place(This is just my personal skill range—there's no need for you to copy it verbatim）

Additionally, I do not recommend enabling tire warmers. Not only does it have no effect on the AI, but it also removes the essence of endurance racing.

Also, I do not recommend fixing the fuel multiplier per track. To be honest, everyone's driving style is vastly different, so there's no one-size-fits-all solution. I have a friend who can only manage 11 laps at Le Mans... Only what suits you is best. For example, at Imola, the AI fuel multiplier setting that achieves perfect synchronization with me is 0.95, not 0.92.

ps:Using the number of laps completed in a single stint as the calibration standard for fuel consumption is inaccurate. Even if both you and the AI pit every 12 laps, differences in remaining fuel can accumulate into an advantage over multiple pit stops. Especially over a long distance—like 24 hours—this will inevitably lead to a different final lap count. Therefore, using the final race lap count as the benchmark is the only accurate approach.

 

[ndslpsp]

One final point concerns the pit-stop windows for fuel and tires. When the AI fuel multiplier is lowered to a certain point, it falls below the tire wear threshold, causing the AI to be forced into a pit stop for tire changes.

For example, in GT3, once tire wear drops below 40%, a mandatory pit stop for tire changes is triggered. At that point, the remaining fuel level only affects refueling time and no longer determines the pit-stop timing.

If you prefer the first style—chasing rather than defending—this is obviously a good approach.

However, tire wear varies greatly across different car models. The AI's Porsche, for instance, seems to have rear tires that wear out first, which staggers the AI's pit-stop timing and greatly increases the randomness of the race.

I set the fuel multiplier very low precisely to let tire wear dictate the AI's race rhythm through pit-stop triggers. For hypercars, since tire changes appear to happen much later, it's probably better to let the fuel window continue to dominate pit-stop strategy instead.

 

[ndslpsp]

As for how to determine the total number of race laps? You don't have to complete the full 6 or 24 hours yourself—you just need to push flat-out for one stint. Before you pit, the window will show that there are 6.4 stints remaining. Multiply that number by the laps you just completed, and you'll get the ideal total lap count for the whole race. Of course, during the race you'll encounter slower cars, get lapped, be blocked, get tangled in battles, and make mistakes, so your actual lap count will definitely fall short of that ideal number. Set that standard as the AI's first-place level, and it will deliver a race that is both exciting and challenging.

 

[Sparviero]

Through the adjustable fuel multiplier, single-player players are presented with only two types of endurance races:

1. Adjust the AI fuel multiplier so that the AI has lower fuel consumption, runs more laps per stint, and stays on track longer. At the same time, lower the difficulty so the AI is slower than you. Your race experience will be: constantly pitting, constantly falling behind, and constantly overtaking again, while doing your best to save fuel and tires to reduce pit stops and run more laps per stint.
2. Adjust the AI fuel multiplier to exactly match the player's consumption (the standard is that at the end of each stint, the AI near your position has almost the same remaining fuel as you). Increase the difficulty so the AI starts off fast, gains time on track early, then slows down at the end of each stint and wastes time with tire changes in the pits. Your race strategy will rely on gaining positions through pit-stop advantages, defending and holding off the AI in the early phase after their pit stops, and pushing hard to chase them down during the late phase of the stint when they slow down.

The calibration of both approaches still ultimately comes down to the hard standard of total race distance—Imola 200 laps for first place(This is just my personal skill range—there's no need for you to copy it verbatim）

Additionally, I do not recommend enabling tire warmers. Not only does it have no effect on the AI, but it also removes the essence of endurance racing.

Also, I do not recommend fixing the fuel multiplier per track. To be honest, everyone's driving style is vastly different, so there's no one-size-fits-all solution. I have a friend who can only manage 11 laps at Le Mans... Only what suits you is best. For example, at Imola, the AI fuel multiplier setting that achieves perfect synchronization with me is 0.95, not 0.92.

ps:Using the number of laps completed in a single stint as the calibration standard for fuel consumption is inaccurate. Even if both you and the AI pit every 12 laps, differences in remaining fuel can accumulate into an advantage over multiple pit stops. Especially over a long distance—like 24 hours—this will inevitably lead to a different final lap count. Therefore, using the final race lap count as the benchmark is the only accurate approach.

Yes, my value are for my style of play but the point is that AI have to pit the same number of times of the player and I'm quite sure that it is not possible setting the same value for each track. However is not really importat matching exactely the same laps per stint among player and AI, but the number of stint per rece lenght. If the final stint is a bit different, doesn't change results too much, expecially in a long race. Obviously if the race is very long (eg 12h), difference in laps per sting may influence the number of total stint, so you have to correct it.

The main advantage with tyre warmer on is to mitigate the damage caused by the AI's more frequent tire changes—something that can't be solved. As you say, the alternative is to turn off tire wear, but I find this approach more realistic, even though it costs the AI about 10 seconds in a 3-hour race and 40 seconds in a 12-hour race.

 

[ndslpsp]

Yes, my value are for my style of play but the point is that AI have to pit the same number of times of the player and I'm quite sure that it is not possible setting the same value for each track. However is not really importat matching exactely the same laps per stint among player and AI, but the number of stint per rece lenght. If the final stint is a bit different, doesn't change results too much, expecially in a long race. Obviously if the race is very long (eg 12h), difference in laps per sting may influence the number of total stint, so you have to correct it.

The main advantage with tyre warmer on is to mitigate the damage caused by the AI's more frequent tire changes—something that can't be solved. As you say, the alternative is to turn off tire wear, but I find this approach more realistic, even though it costs the AI about 10 seconds in a 3-hour race and 40 seconds in a 12-hour race.

No wonder you’d think that… Hypercars really play by a completely different rulebook. GTE, LMP2 and LMP3 also share the same difficulty scaling system as Hypercars. Let me break it down: when I crank the AI strength up to 95% with Hypercars, I get absolutely outclassed. Even the slowest backmarker car can lap me within a single stint. Forget tyre changes during pit stops — I couldn’t catch up even if ai stopped to grab a coffee.Even at 90%, sprint races are an uphill struggle. I’d probably need to drop it down to the 88% just to have a competitive endurance race.

GT3s are a night-and-day contrast, though. At 95% AI strength, I win races with zero contest. I can roll off the pit lane dead last, and surge all the way up to P1 in just half a stint, utterly dominating the entire field.Even with their pit stops for tyre changes, they pull so far clear of me it’s like they’ve disconnected from the race entirely. Even maxed out at 100% difficulty, I pull out a massive lead extremely early—within the first hour of any endurance race, I’m comfortably out front. It just makes the whole race feel totally dull, doesn’t it?

So to keep the competition challenging, I’ve had to reduce the AI’s fuel consumption multiplier. This lets them run longer stints with fewer pit stops, closing the gap and creating a more balanced race against me. Even if I account for my lack of practice with prototype machinery, I strongly suspect the 100% AI setting for GT3s is only equivalent to roughly 92–93% strength for Hypercars.

If you mainly race Hypercars or other prototype vehicles, you won’t need to mess with any of these tweaks. Locking the AI pit windows to match yours is totally logical in that case.

 

[ndslpsp]

I recently ran a Hypercar stint. On my first pit stop, my tires still had about 75–77% remaining, but they'd already developed flat spots and felt pretty awful to drive. The AI's tires were at 60–63% — honestly not that much different, so swapping them was fair enough. Pushing through two stints was really uncomfortable.

In GT3, one set of tires can last four stints, whereas in Hypercar they only last one. That makes the gap between the two classes even bigger. Another important factor is that GT3 isn't as sensitive — even with heavy wear, you can keep driving. But in Hypercar, when the tires are old, it's pure agony. You get trapped in a vicious cycle of constant spinning, overheating, and accelerating wear.

 

[Sparviero]

I recently ran a Hypercar stint. On my first pit stop, my tires still had about 75–77% remaining, but they'd already developed flat spots and felt pretty awful to drive. The AI's tires were at 60–63% — honestly not that much different, so swapping them was fair enough. Pushing through two stints was really uncomfortable.

In GT3, one set of tires can last four stints, whereas in Hypercar they only last one. That makes the gap between the two classes even bigger. Another important factor is that GT3 isn't as sensitive — even with heavy wear, you can keep driving. But in Hypercar, when the tires are old, it's pure agony. You get trapped in a vicious cycle of constant spinning, overheating, and accelerating wear.

I'm not a great driver but I'm quite experience with hypercars and I assure you that hypercars tyres runs comfortably 2 long stints or 3 at Le Mans. Tyres maintain performance up to 30% but the "secrets" are: 1) good introduction of tyres, avoiding flat spot but also rising temperature slowly and uniformly, 2) constant management of energy on tyres (temperature and pressure). If you do that, you can put down your faster times even at 60% and maintain a good pace up to 25%, otherwise your car may become undrivable even at 70%.

This the problem of longer stints by AI due to more frequent tyre change is relevant for hypercars too if you drive them properly.

 

[ndslpsp]

I'm not a great driver but I'm quite experience with hypercars and I assure you that hypercars tyres runs comfortably 2 long stints or 3 at Le Mans. Tyres maintain performance up to 30% but the "secrets" are: 1) good introduction of tyres, avoiding flat spot but also rising temperature slowly and uniformly, 2) constant management of energy on tyres (temperature and pressure). If you do that, you can put down your faster times even at 60% and maintain a good pace up to 25%, otherwise your car may become undrivable even at 70%.

This the problem of longer stints by AI due to more frequent tyre change is relevant for hypercars too if you drive them properly.

You have fantastic tyre management skills. I’ve been playing with tyre wear turned off in arcade mode all this time because of the unfair AI tyre wear mechanics — no rubber degradation, overheating or flat spots whatsoever. It wasn’t until I enabled tyre wear that I realized my old driving style simply doesn’t translate to realistic racing at all.

I’ve found the tyre remaining life percentage doesn’t truly represent actual grip performance. Sometimes the car suddenly becomes unmanageable at around 80% tyre durability, while other times I can still maintain consistent pace even down to 50%. I can’t pinpoint the exact cause; it’s likely hidden internal damage from pushing hard for fast lap times before the tyres hit their optimal operating temperature. I still have a lot to learn when it comes to tyre care.

At the moment, the tyre wear gap between the player and AI is enormous for GT3s. With your level of driving skill, you can run at least four full stints on a single set of tyres, yet the AI changes tyres even more frequently than they do in Hypercar races. The only viable workaround to level the playing field against AI for GT3s is to use unbalanced fuel consumption values to offset this disadvantage. That said, all other racing categories work perfectly fine using the settings you shared.

 

[ndslpsp]

I'm not a great driver but I'm quite experience with hypercars and I assure you that hypercars tyres runs comfortably 2 long stints or 3 at Le Mans. Tyres maintain performance up to 30% but the "secrets" are: 1) good introduction of tyres, avoiding flat spot but also rising temperature slowly and uniformly, 2) constant management of energy on tyres (temperature and pressure). If you do that, you can put down your faster times even at 60% and maintain a good pace up to 25%, otherwise your car may become undrivable even at 70%.

This the problem of longer stints by AI due to more frequent tyre change is relevant for hypercars too if you drive them prop

I did a test yesterday. With the GT3's AI using normal fuel consumption, I could only complete 200 laps at 102% difficulty. After adjusting fuel consumption, I could finish 200 laps at 95% difficulty. Each 2% increase in difficulty adds only one extra lap, and at max 105% difficulty, I can do at most 202 laps. This long-distance capability is just too weak. I think GT3 needs 120% difficulty to save the day—otherwise, the AI has to cheat. As for Hypercar, 90% already makes me despair, and GTE and LMP3 are about the same. The fact that you can handle 94% is already pretty impressive.

 

[ndslpsp]

I'm not a great driver but I'm quite experience with hypercars and I assure you that hypercars tyres runs comfortably 2 long stints or 3 at Le Mans. Tyres maintain performance up to 30% but the "secrets" are: 1) good introduction of tyres, avoiding flat spot but also rising temperature slowly and uniformly, 2) constant management of energy on tyres (temperature and pressure). If you do that, you can put down your faster times even at 60% and maintain a good pace up to 25%, otherwise your car may become undrivable even at 70%.

This the problem of longer stints by AI due to more frequent tyre change is relevant for hypercars too if you drive them properly.

By the way, the tire wear displayed by the AI isn't actually something worth paying close attention to. The tire wear percentage doesn't represent the full remaining capability of the tires. Just as you said—you can handle 30% worn tires with ease, while I struggle once they hit 70%. That simple percentage is not the real gold standard. Only specific lap times and consistent stability are the true metrics for tire management. For a tire management master like you, clearly some real challenges are needed. Turn off fuel consumption, turn on tire wear, and engage in a truly pure tire duel against the AI. You'll see that although their tires appear to degrade quickly, their lap time degradation isn't any worse than yours. I remember at Le Mans, the hypercar managed to last a staggering number of laps—about four stints—before pitting. That's quite terrifying and well worth your challenge...

 

[ndslpsp]

Before we begin this brutal tire endurance battle, there's one more secret I have to share about the fuel multiplier. It doesn't just determine the AI's fuel consumption—it also dictates whether they've got a wooden block under the throttle. During my tests to find the fuel level that best matches my own consumption, I tried values like 0.97, 0.92, and 0.94, but something always felt off during races. No matter what car I drove, my car was always a bit faster on the straights—even without a slipstream, I could gradually catch up to them. And they were always timid in corners, braking early and afraid to go wheel‑to‑wheel. I could always charge from the back of the pack to the front... until I set the value to 2 (I once tried 2.5, but it automatically reverted to 2, so apparently 2 is the maximum). Oh my god, the AI I used to bully became impossible to chase down. They were slippery as eels when overtaking, pouncing on any gap, and absolutely determined in wheel‑to‑wheel battles—whereas before they'd always lift off the throttle. They'd also brake late without hesitation—really tough opponents. Is this the fully unleashed version of the AI? I've already tested it with fuel consumption turned off, and the effect still persists. I think they must have been deliberately saving fuel just to hit the stint lap count you expected. Well... lifting off the throttle just to meet a fuel multiplier, giving up competition—that's pretty boring, isn't it? So let's remove the fuel tank constraints and have a race purely focused on tires. Let me see the AI's true, unfettered strength.

 

[Sparviero]

Before we begin this brutal tire endurance battle, there's one more secret I have to share about the fuel multiplier. It doesn't just determine the AI's fuel consumption—it also dictates whether they've got a wooden block under the throttle. During my tests to find the fuel level that best matches my own consumption, I tried values like 0.97, 0.92, and 0.94, but something always felt off during races. No matter what car I drove, my car was always a bit faster on the straights—even without a slipstream, I could gradually catch up to them. And they were always timid in corners, braking early and afraid to go wheel‑to‑wheel. I could always charge from the back of the pack to the front... until I set the value to 2 (I once tried 2.5, but it automatically reverted to 2, so apparently 2 is the maximum). Oh my god, the AI I used to bully became impossible to chase down. They were slippery as eels when overtaking, pouncing on any gap, and absolutely determined in wheel‑to‑wheel battles—whereas before they'd always lift off the throttle. They'd also brake late without hesitation—really tough opponents. Is this the fully unleashed version of the AI? I've already tested it with fuel consumption turned off, and the effect still persists. I think they must have been deliberately saving fuel just to hit the stint lap count you expected. Well... lifting off the throttle just to meet a fuel multiplier, giving up competition—that's pretty boring, isn't it? So let's remove the fuel tank constraints and have a race purely focused on tires. Let me see the AI's true, unfettered strength.

I'm sorry but I don't understand: do you refer to the general fuel multiplier (2x, 3x, which make all cars consume 2 times fuel) or the fuel rate (the one for AI to set to 0.92 or so to match player fuel consumption)? If it is the latter, the fuel consumption should still be like 2 times the normal, so AI should pit after half of laps of standard setting, or am I wrong?

 

[Sparviero]

By the way, the tire wear displayed by the AI isn't actually something worth paying close attention to. The tire wear percentage doesn't represent the full remaining capability of the tires. Just as you said—you can handle 30% worn tires with ease, while I struggle once they hit 70%. That simple percentage is not the real gold standard. Only specific lap times and consistent stability are the true metrics for tire management. For a tire management master like you, clearly some real challenges are needed. Turn off fuel consumption, turn on tire wear, and engage in a truly pure tire duel against the AI. You'll see that although their tires appear to degrade quickly, their lap time degradation isn't any worse than yours. I remember at Le Mans, the hypercar managed to last a staggering number of laps—about four stints—before pitting. That's quite terrifying and well worth your challenge...

What could works for me is to force the player and AI to change tyres every pit: basically, it can works only if both player/AI are able to do longer stints like 1h30 or 2h. Unfortunatly, it is possible for the AI (eg. setting fuel mult to 0.5-0.7) but not for the player. Racing without fuel consumption at all doesn't works for me, I feel it too much unrealistic and make stints too much longer (I need frequent in-game saving).

 

[ndslpsp]

I'm sorry but I don't understand: do you refer to the general fuel multiplier (2x, 3x, which make all cars consume 2 times fuel) or the fuel rate (the one for AI to set to 0.92 or so to match player fuel consumption)? If it is the latter, the fuel consumption should still be like 2 times the normal, so AI should pit after half of laps of standard setting, or am I wrong?

I'm talking about the fuel multiplier, the 0.92 one. But setting it to 2 causes some cars to never pit—they drag their 0% tires all the way to the end of the race..Could it be that the tire allocation I reserved for them is too little? .
Why set it to 2? Because in the past in rF2, I used to compete with the AI in tire endurance. Since rF2 doesn't allow saves anyway, turning off fuel doesn't matter. They can run 50 laps on softs, and 95–100 laps on hards—pretty impressive.

 

[ndslpsp]

ai brake grip 1.0
ai brake power 1..0
ai corner grip 1.0
ai fuel mult 2.0
Aggressiveness: high
Tire allocation must strictly comply with official standards (including qualifying).
6-hour race: 24 sets
8-hour race: 28 sets
Turn off fuel consumption on the game panel, and turn on tire wear.
If your tire count falls below this standard quota, the AI will keep driving until the tires blow out.