GPT-5.5 tops Artificial Analysis with score of 60

Artificial Analysis ranks GPT-5.5 (xhigh) first on intelligence with a score of 60.

Artificial Analysis has updated its model comparison hub with rankings across intelligence, speed, latency, price, and context window. The site now tracks 523 models, with GPT-5.5 (xhigh) and GPT-5.5 (high) leading the intelligence chart.

What changed

The refreshed Artificial Analysis dashboard lets users compare models by intelligence, output speed, latency, price, context window, and related metrics in one place. Its Intelligence Index v4.0 uses 10 evaluations, including GDPval-AA, Terminal-Bench Hard, SciCode, AA-Omniscience, Humanity's Last Exam, GPQA Diamond, and CritPt.

The current leaderboard puts GPT-5.5 (xhigh) and GPT-5.5 (high) at the top for intelligence, followed by Claude Opus 4.7 (max) and Gemini 3.1 Pro Preview. On the other end of the chart, Mercury 2 leads output speed at 825 tokens per second, Command A+ posts the lowest latency at 0.33 seconds, Qwen3.5 0.8B is the cheapest at $0.01 per million tokens, and Llama 4 Scout offers the largest context window at 10 million tokens.

• 523 models are included in the comparison set.
• Intelligence Index v4.0 uses 10 benchmarks.
• Price is shown as a blended cache-input-output rate.
• Open and proprietary models are both tracked.

Why it matters

For developers, the page makes tradeoffs easier to see: a model can be strong on intelligence but weak on cost, speed, or context length. That matters for agent workflows, RAG systems, coding assistants, and any API choice where token use and latency affect product quality.

It also gives buyers a faster way to compare providers without reading separate benchmark posts. The mix of intelligence, cost, and performance charts can help teams decide whether to optimize for quality, throughput, or budget.

The main takeaway is simple: the best model is no longer just the smartest one, and Artificial Analysis is making that tradeoff visible in a single ranking page.