Skip to main content
Radar

Every change to the models you run, with its source and its date. Releases, price changes, retirements, API changes, and incidents.Every change to the models you run, with its source.

Follow model changes
BenchLM

MMLU-Pro first-party comparison snapshot (MMLU-Pro (Arcee))

We mirror this table; we do not rank on it.

Data verified 36 confirmed releases in the last 30 daysFollow model changes

A display-only MMLU-Pro reference from Arcee AI's Trinity-Large-Thinking launch chart.

Benchmark score on MMLU-Pro (Arcee) — September 18, 2026

We mirror the published score view for MMLU-Pro (Arcee). Claude Opus 4.6 leads the public snapshot at 89.1%, followed by Kimi K2.5 (87.1%) and GLM-5 (85.8%). We do not use these results to rank models overall.

1Proprietary

Claude Opus 4.6

Anthropic

89.1%
Overall 69.21Context 1M
2Open weights

Kimi K2.5

Moonshot AI

87.1%
Overall 54.02Context 256K
3Open weights

GLM-5

Z.AI

85.8%
Overall 61.47Context 200K

6 modelsKnowledgeCurrentDisplay onlyUpdated September 18, 2026

Benchmark score table (6 models)

Score
1
Claude Opus 4.6Anthropic · Closed
89.1%
2
Kimi K2.5Moonshot AI · Open weight
87.1%
3
GLM-5Z.AI · Open weight
85.8%
4
Trinity-Large-ThinkingArcee AI · Open weight
83.4%
5
MiniMax M2.7MiniMax · Open weight
80.8%
6
Trinity-Large-PreviewArcee AI · Open weight
75.2%

The published MMLU-Pro (Arcee) snapshot places Claude Opus 4.6 first at 89.1%. The third row is 3.3 points behind. The broader top-10 range is 13.9 points, so the table still separates the published systems.

6 models have been evaluated on MMLU-Pro (Arcee). The benchmark falls in the Knowledge category. This category carries a 12% weight in BenchLM.ai's overall scoring system. MMLU-Pro (Arcee) is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About MMLU-Pro (Arcee)

Year

2026

Tasks

Professional academic QA

Format

10-way multiple choice

Difficulty

Professional level

BenchLM stores this chart-specific MMLU-Pro row separately so it does not overwrite the standardized weighted MMLU-Pro benchmark values.

Trinity-Large-Thinking: Scaling an Open Source Frontier Agent

Freshness and provenance

Version

MMLU-Pro (Arcee) 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 MMLU-Pro (Arcee) measure?

A display-only MMLU-Pro reference from Arcee AI's Trinity-Large-Thinking launch chart.

Which model scores highest on MMLU-Pro (Arcee)?

Claude Opus 4.6 by Anthropic currently leads with a score of 89.1% on MMLU-Pro (Arcee).

How many models are evaluated on MMLU-Pro (Arcee)?

6 AI models have been evaluated on MMLU-Pro (Arcee) on BenchLM.

Compare top models on MMLU-Pro (Arcee)

Claude Opus 4.6 vs Kimi K2.5Kimi K2.5 vs GLM-5GLM-5 vs Trinity-Large-ThinkingTrinity-Large-Thinking vs MiniMax M2.7
Last updated: September 18, 2026 · BenchLM version MMLU-Pro (Arcee) 2026

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.