Claude Fable 5.2 (High, Default Fallback) vs Claude Haiku 5.5 (Max)
Which is better for the money: Claude Fable 5.2 (High, Default Fallback) by Anthropic or Claude Haiku 5.5 (Max) by Anthropic? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-11-10.
Depends on your budget
Claude Fable 5.2 (High, Default Fallbac… scores 25.3 points higher (61.8 vs 36.5) but costs 21× as much to run the benchmark ($7106 vs $339; flagship vs mid-range cost classes). They are not alternatives at one budget: pick Claude Fable 5.2 (High, Default Fallbac… when you need intelligence above 36.5, and Claude Haiku 5.5 (Max) when 36.5 is enough.
| Metric | Claude Fable 5.2 (High, Default Fallback) | Claude Haiku 5.5 (Max) | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 61.8 | 36.5 | −40.9% Claude Haiku 5.5 (Max) is 40.9% less intelligent |
| Total AA cost (lower is better) | $7,105.77 | $339.47 | −95.2% Claude Haiku 5.5 (Max) is 95.2% cheaper |
| End-to-end time (lower is better) | 36.05s | 371.09s | +929.4% Claude Haiku 5.5 (Max) is 929.4% slower |
| Price per 1M tokens (in / out) | $8.00 / $40.0 | $0.10 / $0.50 | |
| Input modalities | Text, Image | Text, Image |
Claude Haiku 5.5 (Max) vs Claude Fable 5.2 (High, Default Fallback): 40.9% less intelligence, 95.2% less cost to run the AA benchmark, and 929.4% slower end-to-end.
Against the frontier
- Claude Fable 5.2 (High, Default Fallback) is on today's frontier: nothing is both smarter and cheaper.
- Claude Haiku 5.5 (Max) is off the frontier: at the same or lower cost, GPT-6.1 Terra (High) scores 42.40 for $267.
Compare something else
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/claude-fable-5-2-high/claude-haiku-5-5 · Markdown: /api/v1/compare/claude-fable-5-2-high/claude-haiku-5-5?format=md · llms.txt · agent skill