WeirdML (v2) ↗43.71%
Mistral Medium 3.5 · WeirdML · Accuracy · % ↗
Mistral / MODEL EVIDENCE
Mistral Medium 3.5
28 benchmarks · 30 observations · 29 sources · released 2026-04-28
Scores at a glance
28 benchmarks · large dot: Mistral Medium 3.5; small dots: closest models where reported
CritPt ↗0%
Mistral Medium 3.5 · CritPt · Accuracy · % ↗
Surface Evolver Bench ↗26.88%
Mistral Medium 3.5 · Surface Evolver Bench · Mean score · % ↗
ProofBench ↗9 %
Mistral Medium 3.5 · ProofBench · Accuracy ↗
1 more reported result
9% · Mistral Medium 3.5 · ProofBench · Accuracy · % · Epoch AI · ProofBench ↗
Mistral Medium 3.5 · FrontierCode · Main score · % ↗
LMCA ↗26.06%
Mistral Medium 3.5 · LMCA · Score · % ↗
DTBench ↗75.47%
Mistral Medium 3.5 · DTBench · Accuracy · % ↗
SciCode ↗40.16%
Mistral Medium 3.5 · SciCode · Score · % ↗
Text Arena (Coding) ↗1264.92 Elo
Mistral Medium 3.5 · WebDev Arena · Elo ↗
Epoch Capabilities Index (ECI) ↗141.36 ECI
Mistral Medium 3.5 · General ECI · 26 October 2026 fit ↗
Vals Multimodal Index ↗34.77 %
Mistral Medium 3.5 · Vals Multimodal Index · Accuracy ↗
Legal Research Bench ↗9.13 %
Mistral Medium 3.5 · Legal Research Bench · All-Pass ↗
Mistral Medium 3.5 · Harvey's Legal Agent Benchmark v1 · Accuracy ↗
CaseLaw v2 ↗44.16 %
Mistral Medium 3.5 · CaseLaw · Accuracy ↗
Tax Agent Bench ↗4.48 %
Mistral Medium 3.5 · Tax Agent Bench · Accuracy ↗
Vals Finance Agent v2 ↗32.1 %
Mistral Medium 3.5 · Finance Agent v2 · Accuracy ↗
Mistral Medium 3.5 · Excel Modeling Benchmark v1 · Accuracy ↗
CorpFin v2 ↗58.78 %
Mistral Medium 3.5 · CorpFin · Accuracy ↗
MortgageTax & TaxEval v2 ↗28.89 %
Mistral Medium 3.5 · MortgageTax · Accuracy ↗
1 more reported result
67.99 % · Mistral Medium 3.5 · TaxEval · Accuracy · Vals AI · TaxEval ↗
MedCode ↗33.75 %
Mistral Medium 3.5 · MedCode · Accuracy ↗
MedScribe ↗67.73 %
Mistral Medium 3.5 · MedScribe · Accuracy ↗
GPQA Diamond ↗34.85 %
Mistral Medium 3.5 · GPQA Diamond · Accuracy ↗
MMLU-Pro ↗75.33 %
Mistral Medium 3.5 · MMLU Pro · Accuracy ↗
Mistral Medium 3.5 · SAGE · Accuracy ↗
Vibe Code Bench ↗2.89 %
Mistral Medium 3.5 (OpenHands) · Vibe Code Bench · original task set ↗
Code Migration v1.1 ↗5.13 %
Mistral Medium 3.5 · Code Migration v1 · Accuracy ↗
SWE-bench Verified ↗66.4 %
Mistral Medium 3.5 (Mini-SWE-agent) · SWE-bench Verified · Score ↗
ALE-Bench ↗763.98 Performance points
mistral-medium-3.5 · Average contest performance ↗
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
GPT-5.4 Nano →OpenAI
15 of 18 shared benchmark comparisons
Claude Haiku 4.5 →Anthropic
15 of 18 shared benchmark comparisons
15 of 18 shared benchmark comparisons
How similarity is measured
Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
- Arena leaderboard — WebDev evaluation ↗
- SciCode Benchmark | Artificial Analysis ↗
- FrontierCode Leaderboard ↗
- DTBench — Conceptual Reasoning Index ↗
- LMCA — Conceptual Reasoning Index ↗
- Ale Bench — methodology and published results ↗
- CritPt v1 ↗
- AI benchmarks & capabilities ↗
- WeirdML ↗
- CaseLaw — evaluation methodology and results ↗
- Code Migration v1 — evaluation methodology and results ↗
- CorpFin — evaluation methodology and results ↗
- Excel Modeling Benchmark v1 — evaluation methodology and results ↗
- Finance Agent v2 — evaluation methodology and results ↗
- GPQA Diamond — evaluation methodology and results ↗
- Harvey's Legal Agent Benchmark v1 — evaluation methodology and results ↗
- Legal Research Bench — evaluation methodology and results ↗
- MedCode — evaluation methodology and results ↗
- MedScribe — evaluation methodology and results ↗
- MMLU Pro — evaluation methodology and results ↗
- MortgageTax — evaluation methodology and results ↗
- ProofBench — evaluation methodology and results ↗
- SAGE — evaluation methodology and results ↗
- SWE-bench Verified — evaluation methodology and results ↗
- Tax Agent Bench — evaluation methodology and results ↗
- TaxEval — evaluation methodology and results ↗
- Vals Multimodal Index — evaluation methodology and results ↗
- Vibe Code Bench — evaluation methodology and results ↗
- Surface Evolver Bench ↗
Names used in source tables
Mistral Medium 3.5 · Mistral Medium 3.5 (Mini-SWE-agent) · Mistral Medium 3.5 (OpenHands) · mistral-medium-3.5