IFBench (IFBench)

A display-only instruction-following benchmark reference from Arcee AI's Trinity-Large-Thinking comparison chart.

Top Models on IFBench — March 2026

As of March 2026, MiniMax M2.7 leads the IFBench leaderboard with 75.7% , followed by GLM-5 (72.3%) and Kimi K2.5 (70.2%).

5 modelsInstruction FollowingCurrentDisplay onlyUpdated April 1, 2026

According to BenchLM.ai, MiniMax M2.7 leads the IFBench benchmark with a score of 75.7%, followed by GLM-5 (72.3%) and Kimi K2.5 (70.2%). There is significant spread across the leaderboard, making this benchmark effective at differentiating model capabilities.

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

About IFBench

Year

2026

Tasks

Instruction-following tasks

Format

Constraint and compliance evaluation

Difficulty

Instruction precision

BenchLM tracks IFBench as a display-only instruction-following reference when providers publish exact chart values.

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

BenchLM freshness & provenance

Version

IFBench 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.

Leaderboard (5 models)

#1MiniMax M2.7
75.7%
#2GLM-5
72.3%
#3Kimi K2.5
70.2%
#4Claude Opus 4.6
53.1%
#5Trinity-Large-Thinking
52.3%

FAQ

What does IFBench measure?

A display-only instruction-following benchmark reference from Arcee AI's Trinity-Large-Thinking comparison chart.

Which model scores highest on IFBench?

MiniMax M2.7 by MiniMax currently leads with a score of 75.7% on IFBench.

How many models are evaluated on IFBench?

5 AI models have been evaluated on IFBench on BenchLM.

Last updated: April 1, 2026 · BenchLM version IFBench 2026

Weekly LLM Benchmark Digest

Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.

Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.