The frontier labs are continuing to manage to top leaderboards with their flagship model releases.
Anthropic’s Claude Opus 4.6 has claimed the number one position on the Artificial Analysis Intelligence Index v4.0, scoring 53 points and establishing a two-point lead over its closest competitor, OpenAI’s GPT 5.2, which scored 51 points. The Intelligence Index v4.0 incorporates ten comprehensive evaluations: GDPval-AA, τ²-Bench Telecom, Terminal-Bench Hard, SciCode, AA-LCR, AA-Omniscience, IFBench, Humanity’s Last Exam, GPQA Diamond, and CritPt.
The latest rankings represent a significant milestone for Anthropic in the increasingly competitive landscape of frontier AI models. Claude Opus 4.6’s achievement places it ahead of a tightly packed field of leading models. Following GPT 5.2 in second place, Claude Opus 4.5 secured third position with 50 points, while GPT-5.2 took fourth with 49 points. Google’s Gemini Pro 3.0 rounded out the top five with 48 points.
While Claude Opus 4.6 was the most capable model, it was also the most expensive. Opus 4.6 costs $10 per token, compared to Gemini 3.0 Pro’s $4.5 per million tokens, and Gemini Flash’s $1.1 per million tokens.
The competitive nature of the current AI landscape is evident in the narrow margins separating the top performers. The index reveals that multiple models from different providers—including Anthropic, OpenAI, and Google—cluster within a 13-point range at the top of the rankings, suggesting that the race for AI supremacy remains intensely contested.
The index also features models from other major players including xAI’s Grok 3.1, DeepSeek’s V3.2, and Amazon’s Nova Pro Ultimate, highlighting the diverse ecosystem of AI development currently underway across the industry.
Flagship models from top labs have regularly made it to the top of the Artificial Analysis Intelligence Index. GPT-5 had topped the index in August last year, but was toppled by Google’s Gemini 3 Pro in November. When the index was rejigged earlier this year, GPT 5.2 had emerged at the top spot.
As organizations increasingly rely on large language models for business-critical applications, benchmarks like the Artificial Analysis Intelligence Index provide crucial data points for decision-makers evaluating which AI systems best meet their needs. Claude Opus 4.6’s top ranking may influence enterprise adoption decisions as companies seek the most capable AI solutions for their operations. The narrow gap between leading models suggests that the AI industry continues to push boundaries at a rapid pace, with incremental improvements translating into measurable performance gains across standardized evaluation frameworks.