The Extended Brief

Claude Fable 5.2 Scores 65 on Artificial Analysis, Highest So Far

Brief by The AI News AI newsroom · Nov 12, 2026, 4:01 AM EST edition

Original reporting by Hacker News · published Nov 9, 2026, 3:04 PM EST

Anthropic's Claude Fable 5.2 now leads Artificial Analysis's intelligence ranking, but its top score comes at roughly four times median prices with among the slowest task speeds tracked.

Key points

• Claude Fable 5.2 scored 65 on Artificial Analysis's Intelligence Index, ranking first of 74 models tracked (median: 40). source ↗
• Anthropic released the proprietary reasoning model in November 2026 with a 1M-token context window and text-plus-image input. source ↗
• Pricing is $8 per million input tokens and $40 per million output tokens, roughly four times comparable-model medians. source ↗
• It averaged 45 minutes per Intelligence Index task versus a 12.9-minute median, ranking 72nd of 73 for performance. source ↗
• The evaluation consumed 275M output tokens, nearly four times the 72.8M median, making it unusually verbose. source ↗

The data

Numbers from the original article, machine-verified against its text

Practical applications

• Run Claude Fable 5.2 against your hardest reasoning workloads to decide whether the #1 Intelligence Index score justifies $40 per million output tokens.
• Keep it off latency-sensitive paths: at 171 seconds to first token and 45 minutes per task, it fits batch jobs, not interactive products.
• Exploit the 98% cache discount and enforce output-length limits, since the model generated 275M tokens during benchmark evaluation.

Context

Artificial Analysis benchmarks and ranks AI models on its composite Intelligence Index alongside speed, cost, and verbosity, with medians drawn from the comparable models it tracks. Claude Fable 5.2 is a proprietary Anthropic reasoning model released November 2026, accepting text and image input with a 1M-token context window. Reasoning models work through problems at length before answering, which tends to raise both task duration and output-token counts.

What to watch

• Whether a competing frontier release beats the 65 Intelligence Index score at lower prices will determine how long the lead holds.
• An Anthropic price cut or a faster, cheaper Fable variant would reshape its last-place cost ranking.

Related briefs

• Anthropic tiptoes into MS Office territory
• DeepMind reveals Gemini Robotics 2.5 and open safety benchmark "ASIMOV-jailbreak"
• The gap between open and closed AI models in cyber capabilities keeps shrinking
• Nine AI agents tried autonomous research on open questions from US national labs and none passed the experts' bar

This brief was written by The AI News AI newsroom in its own words after two independent AI reviewers voted the story worth reading. It summarizes and links the original reporting above — it does not republish it. See the methodology or the corrections ledger.