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

AIME25 first-party comparison snapshot (AIME25 (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 AIME25 reference from Arcee AI's Trinity-Large-Thinking launch chart.

Benchmark score on AIME25 (Arcee) — September 18, 2026

We mirror the published score view for AIME25 (Arcee). Claude Opus 4.6 leads the public snapshot at 99.8%, followed by Kimi K2.5 (96.3%) and Trinity-Large-Thinking (96.3%). We do not use these results to rank models overall.

1Proprietary

Claude Opus 4.6

Anthropic

99.8%
Overall 69.21Context 1M
2Open weights

Kimi K2.5

Moonshot AI

96.3%
Overall 54.02Context 256K
3Open weights

Trinity-Large-Thinking

Arcee AI

96.3%
Overall 43.27Context 512K

6 modelsMathCurrentDisplay onlyUpdated September 18, 2026

Benchmark score table (6 models)

Score
1
Claude Opus 4.6Anthropic · Closed
99.8%
2
Kimi K2.5Moonshot AI · Open weight
96.3%
3
Trinity-Large-ThinkingArcee AI · Open weight
96.3%
4
GLM-5Z.AI · Open weight
93.3%
5
MiniMax M2.7MiniMax · Open weight
80.0%
6
Trinity-Large-PreviewArcee AI · Open weight
24.0%

The published AIME25 (Arcee) snapshot places Claude Opus 4.6 first at 99.8%. The third row is 3.5 points behind. The broader top-10 range is 75.8 points, so the table still separates the published systems.

6 models have been evaluated on AIME25 (Arcee). The benchmark falls in the Math category. This category carries a 5% weight in BenchLM.ai's overall scoring system. AIME25 (Arcee) is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About AIME25 (Arcee)

Year

2026

Tasks

15 problems

Format

Integer answers 000-999

Difficulty

High school olympiad level

BenchLM stores the Arcee chart version of AIME25 separately so it does not overwrite the weighted AIME 2025 benchmark row.

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

Freshness and provenance

Version

AIME25 (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 AIME25 (Arcee) measure?

A display-only AIME25 reference from Arcee AI's Trinity-Large-Thinking launch chart.

Which model scores highest on AIME25 (Arcee)?

Claude Opus 4.6 by Anthropic currently leads with a score of 99.8% on AIME25 (Arcee).

How many models are evaluated on AIME25 (Arcee)?

6 AI models have been evaluated on AIME25 (Arcee) on BenchLM.

Compare top models on AIME25 (Arcee)

Claude Opus 4.6 vs Kimi K2.5Kimi K2.5 vs Trinity-Large-ThinkingTrinity-Large-Thinking vs GLM-5GLM-5 vs MiniMax M2.7
Last updated: September 18, 2026 · BenchLM version AIME25 (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.