RLTasksby ulam.ai
Models / Anthropic

Anthropic / MODEL EVIDENCE

Claude 3.5 Haiku (Oct 2024)

15 benchmarks · 15 observations · 15 sources · released 2024-10-22

Scores at a glance

15 benchmarks · large dot: Claude 3.5 Haiku (Oct 2024); small dots: closest models where reported

GPQA Diamond ↗38.13%
9.2896.4

Claude 3.5 Haiku (Oct 2024) · GPQA diamond · Best score (across scorers) · % ↗

MATH Level 5 ↗46.36%
3.2998.13

Claude 3.5 Haiku (Oct 2024) · MATH level 5 · Best score (across scorers) · % ↗

BALROG ↗19.3%
3.768.3

Claude 3.5 Haiku (Oct 2024) · Balrog · Average progress · % ↗

CadEval ↗32%
1274

Claude 3.5 Haiku (Oct 2024) · CadEval · Overall pass (%) · % ↗

GeoBench ↗34%
3286

Claude 3.5 Haiku (Oct 2024) · GeoBench · ACW Country % · % ↗

MMLU ↗74.3%
29.587.3

Claude 3.5 Haiku (Oct 2024) · MMLU · EM · % ↗

CritPt ↗0%
032.29

Claude 3.5 Haiku (Oct 2024) · CritPt · Accuracy · % ↗

DTBench ↗56.66%
41.6598.93

Claude 3.5 Haiku (Oct 2024) · DTBench · Accuracy · % ↗

LiveBench ↗43.45%
22.0882.35

Claude 3.5 Haiku (Oct 2024) · LiveBench · Global average · % ↗

SciCode ↗27.43%
1.2766.9

Claude 3.5 Haiku (Oct 2024) · SciCode · Score · % ↗

Closest models by score

Similar positions on shared benchmark leaderboards. Smaller gaps are closer.

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.

Filter comparisons to this model →

Sources

Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.

  1. Aider LLM Leaderboards — Polyglot coding benchmark ↗

    Aider

  2. SciCode Benchmark | Artificial Analysis ↗

    Artificial Analysis

  3. BALROG: Benchmarking Agentic LLM and VLM Reasoning On Games ↗

    Paglieri, Davide, Cupiał, Bartłomiej, Coward, Samuel, Piterbarg, Ulyana, Wolczyk, Maciej, Khan, Akbir et al. · 2024

    All 13 authors

    Paglieri, Davide, Cupiał, Bartłomiej, Coward, Samuel, Piterbarg, Ulyana, Wolczyk, Maciej, Khan, Akbir, Pignatelli, Eduardo, Kuciński, Łukasz, Pinto, Lerrel, Fergus, Rob, Foerster, Jakob Nicolaus, Parker-Holder, Jack, Rocktäschel, Tim

  4. DTBench — Conceptual Reasoning Index ↗

    Conceptual Reasoning Index

  5. Holistic Evaluation of Language Models (HELM) ↗

    Stanford Center for Research on Foundation Models

  6. CritPt v1 ↗

    Epoch AI

  7. FrontierMath Tiers 1-3 (v1, old) ↗

    Epoch AI

  8. GPQA Diamond ↗

    Epoch AI

  9. MATH Level 5 ↗

    Epoch AI

  10. OTIS Mock AIME 2024-2025 ↗

    Epoch AI

  11. GeoBench — GeoGuessr model benchmark ↗

    GeoBench

  12. Writing benchmark — incorporating mandatory story elements ↗

    lechmazur

  13. WeirdML ↗

    Håvard Tveit Ihle

  14. LiveBench — model evaluation leaderboard ↗

    LiveBench Team

  15. CadEval Dashboard — run 20250422_095709 ↗

    willpatrick · 2025

Names used in source tables

Claude 3.5 Haiku (Oct 2024)