Claude Haiku 5.5 (xhigh) vs Claude Haiku 5.5 (max)
Results updated 3 November 2026 · Both in Claude Code · How we benchmark
Verdict
Claude Haiku 5.5 (max) edges Claude Haiku 5.5 (xhigh) by 2 points; Claude Haiku 5.5 (xhigh) is cheaper and faster.
- Cost per task
- $0.017 vs $0.079
- Time per task
- 2.5 min vs 7.8 min (estimated)
- Tokens per attempt
- 458k vs 859k
- Image input
- Yes vs Yes
Claude Haiku 5.5 (max), Claude Code scores higher than Claude Haiku 5.5 (xhigh), Claude Code on all 4 evaluations scored for both. Claude Haiku 5.5 (xhigh), Claude Code is cheaper and faster: $0.017 against $0.079 per task, and 2.5 min against 7.8 min.
Where they differ most
By kind of work, share of tasks passed, 0 to 100 · Higher is better
| Kind of work | Claude Haiku 5.5 (xhigh) | Claude Haiku 5.5 (max) |
|---|---|---|
| GIS analysis | 82 | 91 |
| Climate and time series | 69 | 76 |
| Recognising infeasible requests | 83 | 86 |
| Remote sensing and raster analysis | 76 | 77 |
Where the tasks are set: each kind of work links to its tasks and scoresCharts for each: Claude Haiku 5.5 (xhigh), Claude Code · Claude Haiku 5.5 (max), Claude Code.
Side by side
| Metric | Claude Haiku 5.5 (xhigh), Claude Code | Claude Haiku 5.5 (max), Claude Code |
|---|---|---|
| Geospatial Agent Index: Text only | 76 | 78 |
| Geospatial Agent Index: Text + vision | 77 | 79 |
| Earth-Bench V2 score | 68 | 70 |
| GeoBenchX score | 77 | 78 |
| GS-QA2 score | 81 | 87 |
| Vision agents score | 80 | 81 |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 100 | 97 |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 95 | 100 |
| OGD4All score (standalone; in no index) | 97 | 97 |
| GeoNatureAgent score (standalone; in no index) | 100 | 100 |
| MultiGlobeQA score (standalone; in no index) | 75 | 76 |
| GeoAgentEval score (standalone; in no index) | 92 | 96 |
| Remote sensing and raster analysis score | 76 | 77 |
| GIS analysis score | 82 | 91 |
| Climate and time series score | 69 | 76 |
| Spatial statistics score | 46 | 45 |
| Making maps score | 86 | 85 |
| Recognising infeasible requests score | 83 | 86 |
| Cost per task | $0.017 | $0.079 |
| Time per task (at 100 output tokens/s) | 2.5 min | 7.8 min (estimated) |
| Tokens per attempt | 458k | 859k |
| Turns per attempt | 9.6 | 13.1 |
| Input price per 1M tokens | $0.10 | $0.10 |
| Output price per 1M tokens | $0.50 | $0.50 |
| Context window | 1,000,000 | 1,000,000 |
| Image input | Yes | Yes |
| Developer | Anthropic | Anthropic |
Frequently asked questions
Which scores higher on the Geospatial Agent Index, Claude Haiku 5.5 (xhigh), Claude Code or Claude Haiku 5.5 (max), Claude Code?
Claude Haiku 5.5 (max), Claude Code, 78 to 76.
Which is cheaper per task, Claude Haiku 5.5 (xhigh), Claude Code or Claude Haiku 5.5 (max), Claude Code?
Claude Haiku 5.5 (xhigh), Claude Code, $0.017 to $0.079.
Which writes fewer output tokens per task, Claude Haiku 5.5 (xhigh), Claude Code or Claude Haiku 5.5 (max), Claude Code?
Claude Haiku 5.5 (xhigh), Claude Code, 12k to 43k.
Which has the larger context window, Claude Haiku 5.5 (xhigh), Claude Code or Claude Haiku 5.5 (max), Claude Code?
They are level, at 1,000,000 tokens.