RLTasksby ulam.ai
Models / Meta

Meta / MODEL EVIDENCE

Muse Spark 1.3

47 benchmarks · 78 observations · 48 sources · released 2026-09-02

Scores at a glance

47 benchmarks · large dot: Muse Spark 1.3; small dots: closest models where reported

AA-3D ↗1033.46 Elo
572.551198.83

Muse Spark 1.3 (Max) · AA-3D · Elo ↗

GPQA Diamond ↗94.14%
87.1796.06

Muse Spark 1.3 (Xhigh) · GPQA Diamond · AA implementation ↗

MMMU-Pro ↗82.02%
73.4787.69

Muse Spark 1.3 (Xhigh) · MMMU-Pro · AA implementation ↗

ErdősBench ↗2.860 / 4
2.113.25

Muse Spark 1.3 · ErdősBench · mean progress / 4 · API ↗

CritPt ↗26%
2626

Muse Spark 1.3 (xhigh) · CritPt · Accuracy · % ↗

1 more reported result

24.86% · Muse Spark 1.3 (max) · CritPt · Accuracy · % · Epoch AI · CritPt ↗

GDP.pdf ↗27.6%
235

Muse Spark 1.3 (xhigh) · GDP.pdf · GDP.pdf score · % ↗

WeirdML (v3) ↗12.13%
4.3854.05

Muse Spark 1.3 (xhigh) · WeirdML v3 · Max performance · % ↗

Vals RSI Index ↗0.23 reference score
00.46

Muse Spark 1.3 Max (max) · Compression · reference-anchored score ↗

3 more reported results

0.41 reference score · Muse Spark 1.3 Max (max) · LM training · reference-anchored score · Vals RSI Index · LM training ↗

0.15 reference score · Muse Spark 1.3 Max (max) · Harness engineering · reference-anchored score · Vals RSI Index · Harness engineering ↗

0.00 reference score · Muse Spark 1.3 Max (max) · Post-training · reference-anchored score · Vals RSI Index · Post-training ↗

SAFE-Teen ↗26%
1556

Muse Spark 1.3 (Extra High) · Urgent safety · share of applicable ratings marked fail ↗

2 more reported results

14% · Muse Spark 1.3 (Extra High) · Health advice · share of applicable ratings marked fail · Vals AI · SAFE-Teen ↗

56% · Muse Spark 1.3 (Extra High) · Relationship boundaries · share of applicable ratings marked fail · Vals AI · SAFE-Teen ↗

BTF-3 ↗0.141 Brier scale
0.120.15

Muse Spark 1.3 (xhigh) · Pooled forecasting score ↗

Closest models by score

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

35.4 percentile-point gap

14 of 19 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.

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. Arena leaderboard — WebDev evaluation ↗

    Arena

  2. AA-3D: Interactive 3D Experience Benchmark | Artificial Analysis ↗

    Artificial Analysis

  3. GPQA Diamond Benchmark | Artificial Analysis ↗

    Artificial Analysis

  4. Harvey LAB-AA v1.1 Benchmark | Artificial Analysis ↗

    Artificial Analysis

  5. MMMU-Pro Benchmark | Artificial Analysis ↗

    Artificial Analysis

  6. SciCode Benchmark | Artificial Analysis ↗

    Artificial Analysis

  7. 𝜏³-Banking Benchmark | Artificial Analysis ↗

    Artificial Analysis

  8. Terminal-Bench 2.1 Benchmark | Artificial Analysis ↗

    Artificial Analysis

  9. Cursor · CursorBench ↗

    Cursor

  10. APEX-Agents ↗

    Epoch AI

  11. Chess Puzzles ↗

    Epoch AI

  12. CritPt v1 ↗

    Epoch AI

  13. FrontierMath Tier 4 (v2) ↗

    Epoch AI

  14. FrontierMath Tiers 1-3 (v2) ↗

    Epoch AI

  15. Mystery Game Puzzles ↗

    Epoch AI

  16. OTIS Mock AIME 2024-2025 ↗

    Epoch AI

  17. AI benchmarks & capabilities ↗

    Epoch AI

  18. FutureSearch Evals ↗

    FutureSearch

  19. WeirdML ↗

    Håvard Tveit Ihle

  20. DistressBench — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  21. DrugDiscoveryBench — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  22. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  23. Humanity's Sixth Sense — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  24. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  25. Scale Labs — evaluation methodology and results ↗

    Scale Labs / SEAL · 2026

  26. GDP.pdf Benchmark | Surge AI ↗

    Surge AI

  27. FrontierSWE — software engineering benchmark ↗

    FrontierSWE

  28. ErdősBench: research mathematics, scoped progress and proof audits ↗

    Ulam AI · 2026

  29. Code Migration v1 — evaluation methodology and results ↗

    Vals AI · 2026

  30. CUA-bench — evaluation methodology and results ↗

    Vals AI · 2026

  31. CyberBench — evaluation methodology and results ↗

    Vals AI · 2026

  32. Excel Modeling Benchmark v1 — evaluation methodology and results ↗

    Vals AI · 2026

  33. Finance Agent v2 — evaluation methodology and results ↗

    Vals AI · 2026

  34. Harvey's Legal Agent Benchmark v1 — evaluation methodology and results ↗

    Vals AI · 2026

  35. IOI — evaluation methodology and results ↗

    Vals AI · 2026

  36. Legal Research Bench — evaluation methodology and results ↗

    Vals AI · 2026

  37. SAFE-Teen: Teen Mental Health & Safety | Vals AI A conversation branches after the chatbot turns down a request — evaluation methodology and results ↗

    Vals AI · 2026

  38. MysteryMechanism — evaluation methodology and results ↗

    Vals AI · 2026

  39. ProgramBench — evaluation methodology and results ↗

    Vals AI · 2026

  40. ProofBench — evaluation methodology and results ↗

    Vals AI · 2026

  41. Vals RSI Index — evaluation methodology and results ↗

    Vals AI · 2026

  42. SRE Bench — evaluation methodology and results ↗

    Vals AI · 2026

  43. SWE-bench Multimodal — evaluation methodology and results ↗

    Vals AI · 2026

  44. Tax Agent Bench — evaluation methodology and results ↗

    Vals AI · 2026

  45. Terminal-Bench 4.0 — evaluation methodology and results ↗

    Vals AI · 2026

  46. Terminal-Bench-Science — evaluation methodology and results ↗

    Vals AI · 2026

  47. Vibe Code Bench 1-100 — evaluation methodology and results ↗

    Vals AI · 2026

  48. Vibe Code Bench — evaluation methodology and results ↗

    Vals AI · 2026

Names used in source tables

Muse Spark 1.3 · Muse Spark 1.3 (Extra High) · Muse Spark 1.3 (Max) · Muse Spark 1.3 (Mini-SWE-agent) · Muse Spark 1.3 (OpenHands) · Muse Spark 1.3 (Xhigh) · Muse Spark 1.3 (high) · Muse Spark 1.3 (low) · Muse Spark 1.3 (max) · Muse Spark 1.3 (medium) · Muse Spark 1.3 (xhigh) · Muse Spark 1.3 Max · Muse Spark 1.3 Max (Mini-SWE-agent) · Muse Spark 1.3 Max (OpenHands) · Muse Spark 1.3 Max (max) · Muse-Spark 1.3 (mini-SWE-agent) high · muse-spark-1.3