Mistral Large Instruct 2411 — Benchmarks

Benchmark scores for Mistral Large Instruct 2411 aggregated from public leaderboards (7 boards), with how it ranks among open models and how it compares with its nearest neighbours. See hardware requirements for what you need to run it.

llmrun Composite Score →

82.5

90% range 77.9–87.1 · from 5 boards

CodingNot enough data yet
AgentsNot enough data yet
MathNot enough data yet
ScienceNot enough data yet
Reasoning86.990% range 77–96

Mistral Large Instruct 2411 reasoning benchmarks

Multi-step logic and problem-solving — can the model think through a hard problem rather than recall an answer.

GPQA Diamond
GPQA Diamond is a set of extremely hard, graduate-level science questions (physics, chemistry, biology) written by domain experts and filtered so that skilled non-experts with web access still fail. It measures genuine reasoning rather than memorization.
51.3%
51.3%
median 55.7%open leader 93.2%
Overall best: Gemini 4 Argon (high) 96.4%
46th percentile · #54 of 100 open · #231 of 309 all
#54 of 100
#231 of 309
Independent
Epoch AI · Nov 2024
DTBench
DTBench is a set of handcrafted, expert-validated multiple-choice questions on decision theory — Newcomb-like problems and related puzzles — built by the Conceptual Reasoning Index project with Anthropic. It tests whether a model can follow an abstract argument rather than recall a fact.
60.8%
60.8%
median 68.7%open leader 93.9%
Overall best: Claude Opus 5.5 Max 98.9%
32nd percentile · #48 of 70 open · #190 of 235 all
#48 of 70
#190 of 235
Lab-reported
Epoch AI · Nov 2024
ForecastBench
ForecastBench, from the Forecasting Research Institute, asks models to predict real-world future events, with questions that only resolve after the model's training cutoff so the answers can't have been memorised. It is scored as a difficulty-adjusted Brier index, where higher is better.
61.5
61.5
median 61.5open leader 64.7
Overall best: GPT 5.5 (high) 67.0
48th percentile · #13 of 24 open · #70 of 92 all
#13 of 24
#70 of 92
Lab-reported
Epoch AI · Nov 2024

Mistral Large Instruct 2411 math benchmarks

Symbolic and quantitative problem-solving, from competition math to multi-step calculation.

AIME 2024/2025
AIME (American Invitational Mathematics Examination) is a prestigious high-school competition of hard, integer-answer problems. It's a widely-cited yardstick for multi-step mathematical reasoning, here on the 2024–2025 papers.
7.8%
7.8%
median 54.4%open leader 99.4%
Overall best: Claude Fable 5 (high) 100.0%
27th percentile · #65 of 88 open · #248 of 292 all
#65 of 88
#248 of 292
Independent
Epoch AI · Nov 2024
MATH Level 5
MATH Level 5 covers the hardest tier of competition-mathematics problems, testing multi-step symbolic and quantitative reasoning.
50.3%
50.3%
median 41.6%open leader 96.6%
Overall best: GPT 5 (Aug 07, 2025, high) 98.1%
66th percentile · #14 of 39 open · #62 of 107 all
#14 of 39
#62 of 107
Independent
Epoch AI · Nov 2024

Mistral Large Instruct 2411 knowledge benchmarks

Breadth and depth of factual knowledge across many subjects, paired with the reasoning to apply it.

MMLU-Pro
MMLU-Pro is a harder, cleaned-up successor to MMLU with ten answer choices and more reasoning-heavy questions across 14 subjects, measuring broad knowledge and reasoning together.
67.9%
67.9%
median 52.0%open leader 88.0%
Overall best: Gemini-3.1-Pro 91.2%
72nd percentile · #42 of 146 open · #108 of 257 all
#42 of 146
#108 of 257
Independent
TIGER-Lab

Mistral Large Instruct 2411 instruction following benchmarks

How precisely a model obeys explicit, verifiable instructions — format, length, and content constraints — and, more broadly, how much people prefer its responses head-to-head in blind human voting.

LMArena Text
Arena (arena.ai, formerly LMArena / Chatbot Arena) ranks models by millions of blind, head-to-head human votes on real prompts. The Arena Score shown here is a Bradley-Terry rating computed with Arena's default style control, so higher means people preferred that model's answers more often — it's a relative rating, not a percentage, so compare it only to other models on this board.
1305.6
1305.6
median 1321.4open leader 1488.2
Overall best: Gemini 4 Carbon (High) 1527.5
45th percentile · #104 of 188 open · #274 of 408 all
#104 of 188
#274 of 408
Arena
Arena · Oct 2026

Score vs VRAM in the ≤96 GB tier

Mistral Large Instruct 2411
Other models
Best Score for its VRAM
x: VRAM (GB, log scale) · y: llmrun Score

#40 of 62 that fit ≤96 GB

Each dot is a model at the best quant that fits ≤96 GB. The line steps up to the best Score available at each amount of VRAM; a dot below it is beaten by a model that needs no more memory.

Full ≤96 GB ranking →

Mistral Large Instruct 2411 vs its neighbours

Same boards for the nearest open models per llmrun Score. The highest score in each row is marked.

Board	Mistral Large Instruct 2411this model	Qwen3.8 27Bbest in classQ4_K_M · 18 GB	Qwen2.5 72B Instructnearest by ScoreQ4_K_M · 49 GB	Llama 4 Scout 17B 16E Instructnearest by ScoreQ4_K_M · 77 GB	Mistral Large Instruct 2407nearest by ScoreQ4_K_M · 86 GB
llmrun Score	82.5	122.0	82.2	82.1	81.7
GPQA Diamond	51.3%	—	49.1%	51.8%	49.0%
DTBench	60.8%	88.0%	62.9%	57.9%	61.2%
ForecastBench	61.5	—	61.6	59.1	61.7
AIME 2024/2025	7.8%	—	8.1%	7.8%	8.5%
MATH Level 5	50.3%	—	63.2%	62.3%	44.8%
MMLU-Pro	67.9%	—	71.6%	74.3%	65.9%
LMArena Text	1305.6	1435.6	1302.9	1321.4	1314.4
Compare head to head		Compare	Compare	Compare	Compare

Scores aggregated from public benchmark sources (Epoch AI: 5 boards, Arena: 1 board, TIGER-Lab: 1 board) — each linked from the benchmark pages. llmrun does not run these benchmarks. Percentiles and ranks are among open models in the llmrun catalogue; the median and leader marks refer to the same field.