UnifyBench

RankingEffort benchmarks

Pixtral Large effort benchmarks

Compare measured Low, Medium, High and other settings. Every number keeps its source.

Pixtral Large

9 measured results · 1 reported effort labels · 9 benchmark/harness combinations. Model evidence and sources →

Capability score evidence

Pixtral Large · Non-reasoning: 1.1 overall comparison score. Meets the overall evidence requirements.

Which sources carry the weight?
  • Artificial Analysis · 100.0% direct comparison weight · independent

These shares describe the comparisons entering the fit, not fractions of the final score.

How sensitive is this score?

Remove one evaluator or family and refit the graph against the same overall panel. Values use the matching dataset audited 2026-09-12. These are sensitivity checks, not confidence intervals; ranking eligibility may also change.

Removed evidenceRecomputed scoreChange
Artificial AnalysisComparison path lost—
ARC Prize1.3+0.2
Datacurve0.9-0.2
LiveCodeBench1.0-0.0
Scale1.2+0.1
Terminal-Bench1.1-0.0
Cognition1.1+0.1
LiveBench1.1+0.0
SWE-rebench1.1+0.0
Epoch AI1.1+0.0
Vals AI1.0-0.0
Provider reports · Anthropic1.1+0.0
Provider reports · Moonshot AI1.1-0.0
Provider reports · Meta1.1-0.0
Provider reports · Mistral1.1+0.0
All provider reports1.1-0.0
Every benchmark family
Removed evidenceRecomputed scoreChange
arc-agi1.3+0.2
tau-bench0.9-0.2
deepswe0.9-0.2
aime1.2+0.2
hle1.2+0.1
math5001.2+0.1
ifbench1.0-0.1
omniscience1.0-0.1
frontiercode1.1+0.1
epoch-game-puzzles1.0-0.1
livecodebench1.1+0.1
critpt1.0-0.0
simpleqa1.1+0.0
gpqa1.0-0.0
swe-rebench1.1+0.0
mmmu-pro1.1+0.0
scicode1.1+0.0
multi-swe-bench1.0-0.0
livebench-instructions1.1+0.0
gdp-pdf1.1+0.0
apex-agents1.1+0.0
livebench-language1.1+0.0
frontiermath1.1+0.0
harvey1.0-0.0
aa-lcr1.0-0.0
gmmlu1.1+0.0
swe-atlas-qna1.0-0.0
livebench-coding1.1-0.0
gdpval1.1+0.0
vibe-code1.1-0.0
livebench-reasoning1.1-0.0
arc-agi-31.1-0.0
mmlu-pro1.1-0.0
terminal-bench1.1+0.0
vals-finance-agent1.1-0.0
vals-legal-research1.1-0.0
analyst-agent1.1+0.0
swe-atlas-test-writing1.1-0.0
vals-code-migration1.1-0.0
enterprise-ops1.1+0.0
livebench-data1.1+0.0
vals-excel-modeling1.1-0.0
livebench-math1.1+0.0
swe-atlas-refactoring1.1-0.0
enigma-eval1.1+0.0
briefcase1.1-0.0
itbench1.1-0.0
automationbench1.1+0.0
swe-bench-pro1.1+0.0
What evidence is missing from the fit?

Across all collected settings for this model: 9 observations, 9 contributing, 0 matched without graph weight and 0 excluded. Counts do not establish rank eligibility.

    Inspect every observation and exclusion →

    Capability profile

    One configuration, seven capabilities

    Comparison score · 0–100
    Pixtral Large · Non-reasoning capability radarComparison scores from 0 to 100. Unknown capabilities have no point; lines stop at gaps. Hollow points are preliminary. Exact values and support labels follow the chart.AgenticHard reasoningCodingHuman prefKnowledgeMultimodalLong context50100Pixtral Large · Non-reasoning · Agentic: 4.4 · PreliminaryPixtral Large · Non-reasoning · Hard reasoning: 0.7 · SupportedPixtral Large · Non-reasoning · Coding: 0.6 · PreliminaryPixtral Large · Non-reasoning · Knowledge: 2.1 · SupportedPixtral Large · Non-reasoning · Multimodal: 2.0 · Preliminary

    Gaps are unknown, not zero. Hollow points are preliminary.

    • Agentic

      A4.4Preliminary

    • Hard reasoning

      A0.7Supported

    • Coding

      A0.6Preliminary

    • Human pref

      Unknown

    • Knowledge

      A2.1Supported

    • Multimodal

      A2.0Preliminary

    • Long context

      Unknown

    Each profile belongs to the named configuration. Capability scores use a shared panel for all axes, distinct from the overall ranking panel; they are not accuracy percentages or direct head-to-head wins.

    Compare this configuration →Explore the model evidence profile →

    These are published benchmark scores. Supported effort estimates also appear on the Capability leaderboard. Results from different harnesses occupy different rows. Missing cells stay unknown; no scores are copied between effort levels.

    Benchmark / sourceNon-reasoning
    AIME 2025Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings
    Benchmark details

    Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis

    Unit: percent. Source collected 2026-09-12.

    2.33%
    Reported settings

    Pixtral Large; Reported effort: non-reasoning; Model source: https://artificialanalysis.ai/models/pixtral-large-2411; Mode from evaluator isReasoning metadata; effort budget not specified

    GPQA DiamondArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings
    Benchmark details

    Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis

    Unit: percent. Source collected 2026-09-12.

    50.51%
    Reported settings

    Pixtral Large; Reported effort: non-reasoning; Model source: https://artificialanalysis.ai/models/pixtral-large-2411; Mode from evaluator isReasoning metadata; effort budget not specified

    Humanity's Last ExamArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings
    Benchmark details

    Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis

    Unit: percent. Source collected 2026-09-12.

    2.8%
    Reported settings

    Pixtral Large; Reported effort: non-reasoning; Model source: https://artificialanalysis.ai/models/pixtral-large-2411; Mode from evaluator isReasoning metadata; effort budget not specified

    IFBenchArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings
    Benchmark details

    Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis

    Unit: percent. Source collected 2026-09-12.

    34.49%
    Reported settings

    Pixtral Large; Reported effort: non-reasoning; Model source: https://artificialanalysis.ai/models/pixtral-large-2411; Mode from evaluator isReasoning metadata; effort budget not specified

    LiveCodeBenchArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings
    Benchmark details

    Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis

    Unit: percent. Source collected 2026-09-12.

    26.14%
    Reported settings

    Pixtral Large; Reported effort: non-reasoning; Model source: https://artificialanalysis.ai/models/pixtral-large-2411; Mode from evaluator isReasoning metadata; effort budget not specified

    MATH-500Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings
    Benchmark details

    Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis

    Unit: percent. Source collected 2026-09-12.

    71.4%
    Reported settings

    Pixtral Large; Reported effort: non-reasoning; Model source: https://artificialanalysis.ai/models/pixtral-large-2411; Mode from evaluator isReasoning metadata; effort budget not specified

    MMLU-ProArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings
    Benchmark details

    Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis

    Unit: percent. Source collected 2026-09-12.

    70.11%
    Reported settings

    Pixtral Large; Reported effort: non-reasoning; Model source: https://artificialanalysis.ai/models/pixtral-large-2411; Mode from evaluator isReasoning metadata; effort budget not specified

    MMMU-ProArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings
    Benchmark details

    Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis

    Unit: percent. Source collected 2026-09-12.

    50.58%
    Reported settings

    Pixtral Large; Reported effort: non-reasoning; Model source: https://artificialanalysis.ai/models/pixtral-large-2411; Mode from evaluator isReasoning metadata; effort budget not specified

    𝜏²-Bench TelecomArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings
    Benchmark details

    Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis

    Unit: percent. Source collected 2026-09-12.

    36.55%
    Reported settings

    Pixtral Large; Reported effort: non-reasoning; Model source: https://artificialanalysis.ai/models/pixtral-large-2411; Mode from evaluator isReasoning metadata; effort budget not specified

    What these comparisons mean

    Each source reports its own effort labels. Matching labels do not establish equal compute budgets or identical fallback behavior. Open Reported settings to inspect the exact model name, agent and task coverage. Multiple results at one level remain visible rather than selecting the best. Source-reported None or Non-reasoning labels are not treated as evidence that an API supports those settings.

    These measurements use their own dated source collection. The leaderboard fits identified configurations jointly; it contains no pooled model ratings. Explicit thinking budgets are separate configurations. Read the methodology.

    Download all effort benchmark results · 7,424 results across 240 models · Collected 2026-09-12