Noometry

Model comparison

GPT-5.4 vs Mistral Large

GPT-5.4 is the stronger model overall, scoring 59.4 to 31.9 on the Noometry Index. Mistral Large costs 1.9× less per token, which makes it the better buy when GPT-5.4's lead doesn't matter for your workload.

Last verified . 29 shared benchmarks.

GPT-5.4 OpenAI

59.4

Rank #16 Confirmed

Mistral Large Mistral AI

31.9

Rank #263 Confirmed

Summary

  • They share 29 benchmarks with published results for both. GPT-5.4 scores higher in 9 categories and Mistral Large in 0 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in math, where GPT-5.4 leads 73.5 to 18.2.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 97.8% for GPT-5.4 and 8.5% for Mistral Large.
  • Mistral Large is cheaper at $2 / $6 per million input/output tokens, against $2.50 / $15 for GPT-5.4.
  • GPT-5.4 accepts more context: 1.05M tokens versus 131K.
  • Mistral Large has downloadable open weights; the other is API-only.

Side by side

GPT-5.4Mistral Large
ProviderOpenAIMistral AI
Noometry Index59.431.9
Released2026-03-052024-02-26
WeightsProprietaryOpen
Context window1.05M131K
Max output128K16K
Input $ / M tokens$2.50$2
Output $ / M tokens$15$6
Results tracked6851

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding GPT-5.4 leads

GPT-5.4: 52.6 (#33), Mistral Large: 34.3 (#240)

BenchmarkGPT-5.4Mistral Large
SciCode56.6%36.2%
LMArena Coding14971277
ALE-Bench1,607264.7
SWE-bench Verified76.9%—
DeepSWE51.8%—
LMArena WebDev1465—
GSO31.4%—
WeirdML77.7%—
BigCodeBench Instruct—30%
LiveBench Coding—47.1%
MirrorCode15.6%—
BigCodeBench Complete—38.3%
AlgoTune1.85—
HumanEval+—62.2%
MBPP+—59.5%

Agentic & Tool Use GPT-5.4 leads

GPT-5.4: 46.5 (#13), Mistral Large: 28.6 (#89)

BenchmarkGPT-5.4Mistral Large
Terminal-Bench81.8%—
APEX-Agents52.4%—
Berkeley Function Calling Leaderboard—38.4%
τ²-bench Banking39.4%—
DeepResearch Bench35.1%—
PostTrainBench19%—
GBAEval45.1%—
LMArena Search1197—
METR Time Horizons74.3%—
Vending-Bench 26,144—

Reasoning GPT-5.4 leads

GPT-5.4: 61.8 (#19), Mistral Large: 15.8 (#310)

BenchmarkGPT-5.4Mistral Large
CritPt23.4%0%
LMArena Hard Prompts14851257
DTBench94.4%65.1%
LMCA52%16.7%
Epoch Capabilities Index156.81128.52
ForecastBench59.557.1
ARC-AGI-274%—
SimpleBench—22.5%
Kagi LLM Benchmark63.8%—
NYT Connections (extended)91.3%—
ARC-AGI-193.7%—
Chess Puzzles44%—
EnigmaEval16%—
Thematic Generalization80%—
EBR-Bench25.4%—
LiveBench Reasoning—43.5%
Mystery Game Puzzles37%—
LiveBench Data Analysis—50.1%
LiveBench—48.4%

Math GPT-5.4 leads

GPT-5.4: 73.5 (#19), Mistral Large: 18.2 (#291)

BenchmarkGPT-5.4Mistral Large
OTIS Mock AIME 2024-202597.8%8.5%
LMArena Math14881262
FrontierMath (Feb 2025 set)47.6%0.3%
FrontierMath (Tiers 1-3)78.6%—
FrontierMath Tier 449%—
MathArena Final-Answer Competitions83.1%—
ProofBench56%—
Omni-MATH—28.1%
LiveBench Math—42.5%
MATH Level 5—50.3%
FrontierMath Tier 4 (v1)27.1%—

Knowledge GPT-5.4 leads

GPT-5.4: 65.3 (#14), Mistral Large: 30.1 (#230)

BenchmarkGPT-5.4Mistral Large
GPQA Diamond93.3%51.3%
Vectara Hallucination Rate7%4.5%
LMArena Expert15071232
Humanity's Last Exam36.2%—
SimpleQA Verified45.1%—
MMLU-Pro—59.9%
Confabulations—21.4%
GPQA (HELM)—43.5%
MMLU—80%

Multimodal Not comparable

GPT-5.4: 43.7 (#20), Mistral Large: —

BenchmarkGPT-5.4Mistral Large
LMArena Vision1303—
Blueprint-Bench 227.1%—
Furniture Assembly37.5%—
LMArena Document1471—

Multilingual GPT-5.4 leads

GPT-5.4: 56.2 (#23), Mistral Large: 40.0 (#219)

BenchmarkGPT-5.4Mistral Large
LMArena Non-English14651237
LMArena Chinese15191240
LMArena French14931325
LMArena German14721254
LMArena Japanese14851188
LMArena Korean14481202
LMArena Russian14801257
LMArena Spanish14541268

Instruction Following GPT-5.4 leads

GPT-5.4: 77.1 (#27), Mistral Large: 67.9 (#191)

BenchmarkGPT-5.4Mistral Large
LMArena Instruction Following14691249
LiveBench Instruction Following—67.9%
IFEval—87.7%

Long Context GPT-5.4 leads

GPT-5.4: 50.3 (#8), Mistral Large: 38.3 (#199)

BenchmarkGPT-5.4Mistral Large
LMArena Longer Query14731261
CL-bench27.9%—
CL-bench Life21.7%—

Writing & Preference GPT-5.4 leads

GPT-5.4: 71.9 (#17), Mistral Large: 40.7 (#242)

BenchmarkGPT-5.4Mistral Large
LMArena Text14691266
LMArena Creative Writing14391243
EQ-Bench Creative Writing1840985
LMArena Multi-Turn14821260
Short-Story Creative Writing—69%
WildBench—80.1%
EQ-Bench 41272—
LiveBench Language—39.4%

Frequently asked questions

Is GPT-5.4 better than Mistral Large?

GPT-5.4 is the stronger model overall, scoring 59.4 to 31.9 on the Noometry Index. Mistral Large costs 1.9× less per token, which makes it the better buy when GPT-5.4's lead doesn't matter for your workload.

Which is cheaper, GPT-5.4 or Mistral Large?

Mistral Large is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-5.4 lists at $2.50 and $15.

Is GPT-5.4 or Mistral Large better for coding?

GPT-5.4 scores higher on coding benchmarks: 52.6 versus 34.3 in the Noometry coding category.

Which has the bigger context window?

GPT-5.4 does, with 1.05M tokens against 131K.

How many benchmarks do GPT-5.4 and Mistral Large share?

29 benchmarks have published results for both models. GPT-5.4 has 68 scored results on Noometry and Mistral Large has 51.

Related comparisons

Go deeper