Skip to main content
Radar

Five or fewer confirmed AI changes, with original sources, on mornings when something changed.A free source-linked morning brief.

Start free brief
BenchLM
Advertise

Benchmark profile

InferenceEval

An internal scientific-inference evaluation reported in the Grok 4.6 model card.

Data verified 26 confirmed releases in the last 30 daysStart free brief

Benchmark score on InferenceEval — August 12, 2026

BenchLM mirrors the published score view for InferenceEval. Grok 4.6 leads the public snapshot at 46.9% , followed by GPT-5.6 Sol (44.1%) and Grok 4.5 (41.3%). We do not use these results to rank models overall.

1Closed

Grok 4.6

xAI

grok-4-6

46.9%
Overall 62.44Context 500K
2Closed

GPT-5.6 Sol

OpenAI

gpt-5-6-sol

44.1%
Overall 81.23Context 1.05M
3Closed

Grok 4.5

xAI

grok-4-5

41.3%
Overall 75.42Context 500K
6 modelsCodingCurrentDisplay onlyUpdated August 12, 2026

Benchmark score table (6 models)

Score
1
Grok 4.6xAI · Closed
46.9%
2
GPT-5.6 SolOpenAI · Closed
44.1%
3
Grok 4.5xAI · Closed
41.3%
4
Claude Opus 4.8Anthropic · Closed
40.7%
5
Kimi K3Moonshot AI · Closed
39.8%
6
GPT-5.5OpenAI · Closed
38.9%

The published InferenceEval snapshot places Grok 4.6 first at 46.9%. The third row is 5.6 points behind. The broader top-10 range is 8.0 points, so many of the published results sit in a relatively narrow band.

6 models have been evaluated on InferenceEval. The benchmark falls in the Coding category. This category carries a 20% weight in BenchLM.ai's overall scoring system. InferenceEval is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About InferenceEval

Year

2026

Tasks

Internal scientific-inference tasks

Format

Accuracy

Difficulty

Frontier scientific reasoning

No matching public benchmark repository or result page was found. BenchLM keeps the card's exact comparison rows display-only.

Grok 4.6 Model Card

BenchLM freshness & provenance

Version

InferenceEval 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.

FAQ

What does InferenceEval measure?

An internal scientific-inference evaluation reported in the Grok 4.6 model card.

Which model scores highest on InferenceEval?

Grok 4.6 by xAI currently leads with a score of 46.9% on InferenceEval.

How many models are evaluated on InferenceEval?

6 AI models have been evaluated on InferenceEval on BenchLM.

Compare Top Models on InferenceEval

Grok 4.6 vs GPT-5.6 SolGPT-5.6 Sol vs Grok 4.5Grok 4.5 vs Claude Opus 4.8Claude Opus 4.8 vs Kimi K3
Last updated: August 12, 2026 · BenchLM version InferenceEval 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.