Scale Labs MultiChallenge (MultiChallenge)

We show this table for reference; we do not rank on it.

A Scale Labs public leaderboard mirrored as display-only reference data. It does not affect BenchLM rankings.

Scale score on MultiChallenge — September 29, 2026 snapshot

We mirror the published scale score view for MultiChallenge. Muse Spark leads the public snapshot at 75.5%, followed by Muse Spark 1.1\n (75.3%) and gemini-3.1-pro-preview (71.4%). We do not use these results to rank models overall.

1Proprietary

Muse Spark

Scale score table (30 models)

Score

Muse SparkMeta · Closed

75.5%

Muse Spark 1.1\nMeta

75.3%

gemini-3.1-pro-previewGoogle

71.4%

GPT-5.4 ProOpenAI · Closed

69.2%

gemini-3-pro-previewGoogle

65.7%

gpt-5.1-2025-11-13-thinkingOpenAI

63.4%

gpt-5-thinkingOpenAI

63.2%

o3-pro-2025-06-10-reasoning-highOpenAI

62.4%

Kimi K2.5Moonshot AI · Open weight

61.4%

gemini-3.1-flash-lite-previewGoogle

60.6%

gpt-5-mini-thinkingOpenAI

59.0%

Claude Opus 4.5 ThinkingAnthropic · Closed

59.0%

claude-4-opus-thinkingAnthropic

58.6%

claude-opus-4-1-20250805-thinkingAnthropic

57.2%

claude-4-sonnet-thinkingAnthropic

57.1%

o3-2025-04-16-reasoning-highOpenAI

56.6%

claude-opus-4-6 (Non-Thinking)Anthropic

56.0%

kimi-k2-thinkingMoonshot AI

55.4%

Claude Sonnet 4.5 ThinkingAnthropic · Closed

55.3%

Gemini 2.5 ProGoogle · Closed

53.6%

claude-3-7-sonnet-thinkingAnthropic

51.6%

gpt-5.1-2025-11-13-instantOpenAI

51.2%

Claude Haiku 4.5 ThinkingAnthropic · Closed

50.5%

deepseek-v3p1Deepseek

46.1%

GPT-OSS 120BOpenAI · Open weight

45.3%

o4-mini-2025-04-16-reasoning-highOpenAI

44.9%

qwen3-235b-a22bAlibaba

41.2%

GPT-4.1OpenAI · Closed

39.4%

claude-opus-4-6-thinking-maxAnthropic

37.1%

gemini-2-0-flashGoogle

36.4%

How to read this leaderboard

Compare the published configurations as complete evaluation systems. The source can combine a base model, agent scaffold, tools, budget, and inference setting in each result.

Operator receipt: 30 sourced rows are currently displayable on this page; the leading published row is Muse Spark at 75.5%.

Honest limit: This Scale table is display-only context, not benchmark provenance or a weighted model-only comparison.

MultiChallenge on Scale Labs

How BenchLM shows MultiChallenge

BenchLM mirrors 30 published rows from Scale Labs’ public MultiChallenge leaderboard, captured on September 29, 2026 snapshot.

The table is display only. It is useful context for a published agent or model configuration, but it does not enter BenchLM’s overall or category rankings.

Snapshot

30 published rowsScale Labs sourceDisplay only

MultiChallenge on Scale Labs

The published MultiChallenge snapshot places Muse Spark first at 75.5%. The third row is 4.1 points behind. The broader top-10 range is 14.9 points, so the table still separates the published systems.

30 models have been evaluated on MultiChallenge. The benchmark falls in the Instruction Following category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. MultiChallenge is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About MultiChallenge

Year

2026

Tasks

30 published rows

Format

Published Scale leaderboard score

Difficulty

External agent and model evaluation

BenchLM mirrors 30 published rows from the MultiChallenge public table captured on September 29, 2026 snapshot.

Scale Labs leaderboard Public benchmark source

Freshness and provenance

Version

MultiChallenge 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does MultiChallenge measure?

A Scale Labs public leaderboard mirrored as display-only reference data. It does not affect BenchLM rankings.

Which model leads the published MultiChallenge snapshot?

Muse Spark currently leads the published MultiChallenge snapshot with 75.5% scale score. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on MultiChallenge?

The September 29, 2026 snapshot snapshot contains 30 AI models.

Last updated: September 29, 2026 snapshot · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 2,000+ readers.

One email each week. Unsubscribe anytime.