MetaWatchby Arkwright
RankingsToolsLatestAbout
RankingsToolsLatestAbout
MetaWatchby Arkwright

MetaWatch is a publication of Arkwright Solutions. Signal from noise in AI tooling.

Content

RankingsToolsArticlesRSS Feed

Company

AboutArkwright Solutions
© 2026 Arkwright Solutions. All rights reserved.
← All Rankings

Best Models

Frontier text models for general reasoning and coding

1

Claude Opus 4.5

80MetaScore

Most capable Claude model. Breakthrough reasoning and agentic capabilities.

high
Tool profileLMArenaArtificial AnalysisExpert consensus
2

GPT-5.2

76MetaScore

OpenAI's latest frontier. Top of Artificial Analysis rankings.

high
Tool profileLMArenaArtificial AnalysisExpert consensus
3

DeepSeek R1

75MetaScore

Open weights reasoning model rivaling o1. Exceptional value.

high
Tool profileLMArenaArtificial AnalysisExpert consensus

Methodology

Expert Sentiment:55%
Leaderboards:35%
Value Score:10%

Expert sentiment sourced from curated X accounts via Grok. Leaderboard data from LMArena, Artificial Analysis, and others.