Claude Haiku 5.5 (Xhigh) vs Gemini 4 Carbon (Low)
Which is better for the money: Claude Haiku 5.5 (Xhigh) by Anthropic or Gemini 4 Carbon (Low) by Google? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-11-04.
Trade-off
Verdict
Depends on your budget
Gemini 4 Carbon (Low) scores 19.6 points higher (54.2 vs 34.6) but costs 30× as much to run the benchmark ($5725 vs $193; flagship vs mid-range cost classes). They are not alternatives at one budget: pick Gemini 4 Carbon (Low) when you need intelligence above 34.6, and Claude Haiku 5.5 (Xhigh) when 34.6 is enough.
| Metric | Claude Haiku 5.5 (Xhigh) | Gemini 4 Carbon (Low) | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 34.6 | 54.2 | +56.6% Gemini 4 Carbon (Low) is 56.6% more intelligent |
| Total AA cost (lower is better) | $192.70 | $5,725.02 | +2870.9% Gemini 4 Carbon (Low) is 30× more expensive |
| End-to-end time (lower is better) | 66.84s | 16.28s | −75.6% Gemini 4 Carbon (Low) is 75.6% faster |
| Price per 1M tokens (in / out) | $0.10 / $0.50 | $4.00 / $20.0 | |
| Input modalities | Text, Image | Text, Image, Speech, Video |
Gemini 4 Carbon (Low) vs Claude Haiku 5.5 (Xhigh): 56.6% more intelligence, 2870.9% more cost to run the AA benchmark, and 75.6% faster end-to-end.
Today's Pareto frontier
Against the frontier
- Claude Haiku 5.5 (Xhigh) is off the frontier: at the same or lower cost, GPT-6.1 Terra (Medium) scores 38.20 for $190.
- Gemini 4 Carbon (Low) is off the frontier: at the same or lower cost, Claude Opus 5.5 (Xhigh, Default Fallback) scores 57.40 for $4766.
Claude Haiku 5.5 (Xhigh)Gemini 4 Carbon (Low)Today's Pareto frontierEvery other model
Matchup
Compare something else
Agents:
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/claude-haiku-5-5-xhigh/gemini-4-carbon-low · Markdown: /api/v1/compare/claude-haiku-5-5-xhigh/gemini-4-carbon-low?format=md · llms.txt · agent skill