Magistral Small 2506 — Benchmarks

Benchmark scores for Magistral Small 2506 aggregated from public leaderboards (4 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 →

88.7

90% range 84.3–93.1 · from 4 boards

CodingNot enough data yet
AgentsNot enough data yet
MathNot enough data yet
ScienceNot enough data yet
Reasoning88.190% range 72–104

Magistral Small 2506 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.
56.1%
56.1%
median 54.0%open leader 93.1%
Overall best: Gemini 4 Argon (high) 96.4%
53rd percentile · #47 of 98 open · #213 of 300 all
#47 of 98
#213 of 300
Independent
Epoch AI · Jun 2025
ARC-AGI-2
ARC-AGI-2 is the second-generation ARC Prize puzzle set, rebuilt so that tasks solved by brute-force search or memorization were removed and every problem still requires inferring a novel rule from a handful of examples. It is markedly harder than the original ARC-AGI, and most models — open and closed alike — score near zero.
0.0%
0.0%
median 11.8%open leader 72.9%
Overall best: GPT 6 Astra Max 95.0%
lowest · #19 of 19 open · #221 of 222 all
#19 of 19
#221 of 222
Independent
Epoch AI · Jun 2025
ARC-AGI
ARC-AGI tests fluid, abstract reasoning on small visual grid puzzles where each task follows a novel rule the model must infer from a few examples. It deliberately resists memorization and is one of the most-watched measures of general reasoning progress.
5.0%
5.0%
median 65.3%open leader 94.5%
Overall best: Claude Fable 5 (xhigh) 98.5%
11th percentile · #17 of 19 open · #225 of 232 all
#17 of 19
#225 of 232
Independent
Epoch AI · Jun 2025

Magistral Small 2506 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.
30.0%
30.0%
median 52.8%open leader 98.6%
Overall best: Claude Fable 5 (high) 100.0%
42nd percentile · #50 of 86 open · #216 of 283 all
#50 of 86
#216 of 283
Independent
Epoch AI · Jun 2025

Score vs VRAM in the 24 GB tier

Magistral Small 2506
Other models
Best Score for its VRAM
x: VRAM (GB) · y: llmrun Score

#19 of 35 that fit 24 GB

Each dot is a model at the best quant that fits a 24 GB GPU. The line joins the models with the best Score for their VRAM; a dot below it is beaten by a model that needs no more memory.

Full 24 GB ranking →

Magistral Small 2506 vs its neighbours

Same boards for the nearest open models per llmrun Score and the best model that fits 24 GB. The highest score in each row is marked.

Board	Magistral Small 2506this model	Qwen3.8 27Bbest in classQ4_K_M · 18 GB	Mistral Small 3.2 24B Instruct 2506nearest by ScoreQ4_K_M · 18 GB	Magistral Small 2509nearest by ScoreQ4_K_M · 18 GB	Phi 4nearest by ScoreQ4_K_M · 13 GB
llmrun Score	88.7	122.1	85.7	85.6	83.9
GPQA Diamond	56.1%	—	49.0%	47.6%	56.1%
ARC-AGI-2	0.0%	42.4%	—	—	—
ARC-AGI	5.0%	87.5%	—	—	—
AIME 2024/2025	30.0%	—	30.3%	28.1%	13.8%
Compare head to head		Compare	Compare	Compare	Compare

Scores aggregated from public benchmark sources (Epoch AI: 4 boards) — 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.