Claude Opus 5.5 (max) vs GPT-6.1 Sol (medium)
Results updated 5 November 2026 · In Claude Code and Codex, so part of any gap can come from the harness · How we benchmark
Verdict
GPT-6.1 Sol (medium) edges Claude Opus 5.5 (max) by 2 points, and is cheaper and faster.
- Cost per task
- $1.00 vs $0.046
- Time per task
- 5.3 min (estimated) vs 0.4 min
- Tokens per attempt
- 674k vs 89k
- Image input
- Yes vs Yes
Claude Opus 5.5 (max), Claude Code scores higher on Earth-Bench V2; GPT-6.1 Sol (medium), Codex scores higher on GeoBenchX and GS-QA2. GPT-6.1 Sol (medium), Codex is cheaper and faster: $0.046 against $1.00 per task, and 0.4 min against 5.3 min.
Where they differ most
By kind of work, share of tasks passed, 0 to 100 · Higher is better
| Kind of work | Claude Opus 5.5 (max) | GPT-6.1 Sol (medium) |
|---|---|---|
| Spatial statistics | 57 | 31 |
| Recognising infeasible requests | 73 | 90 |
| Remote sensing and raster analysis | 87 | 83 |
| Making maps | 77 | 81 |
Where the tasks are set: each kind of work links to its tasks and scoresCharts for each: Claude Opus 5.5 (max), Claude Code · GPT-6.1 Sol (medium), Codex.
Side by side
| Metric | Claude Opus 5.5 (max), Claude Code | GPT-6.1 Sol (medium), Codex |
|---|---|---|
| Geospatial Agent Index: Text only | 80 | 82 |
| Geospatial Agent Index: Text + vision | 83 | No data |
| Earth-Bench V2 score | 80 | 72 |
| GeoBenchX score | 70 | 79 |
| GS-QA2 score | 92 | 93 |
| Vision agents score | 92 | No data |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 100 | 100 |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 100 | 100 |
| OGD4All score (standalone; in no index) | 95 | No data |
| GeoNatureAgent score (standalone; in no index) | 100 | No data |
| MultiGlobeQA score (standalone; in no index) | 76 | No data |
| GeoAgentEval score (standalone; in no index) | 100 | No data |
| Remote sensing and raster analysis score | 87 | 83 |
| GIS analysis score | 92 | 94 |
| Climate and time series score | 74 | 75 |
| Spatial statistics score | 57 | 31 |
| Making maps score | 77 | 81 |
| Recognising infeasible requests score | 73 | 90 |
| Cost per task | $1.00 | $0.046 |
| Time per task (at 100 output tokens/s) | 5.3 min (estimated) | 0.4 min |
| Tokens per attempt | 674k | 89k |
| Turns per attempt | 13.1 | 5.8 |
| Input price per 1M tokens | $4.00 | $2.00 |
| Output price per 1M tokens | $20.00 | $10.00 |
| Context window | 1,000,000 | 1,050,000 |
| Image input | Yes | Yes |
| Developer | Anthropic | OpenAI |
Frequently asked questions
Which scores higher on the Geospatial Agent Index, Claude Opus 5.5 (max), Claude Code or GPT-6.1 Sol (medium), Codex?
GPT-6.1 Sol (medium), Codex, 82 to 80.
Which is cheaper per task, Claude Opus 5.5 (max), Claude Code or GPT-6.1 Sol (medium), Codex?
GPT-6.1 Sol (medium), Codex, $0.046 to $1.00.
Which writes fewer output tokens per task, Claude Opus 5.5 (max), Claude Code or GPT-6.1 Sol (medium), Codex?
GPT-6.1 Sol (medium), Codex, 1k to 26k.
Which has the larger context window, Claude Opus 5.5 (max), Claude Code or GPT-6.1 Sol (medium), Codex?
GPT-6.1 Sol (medium), Codex, 1,050,000 tokens to 1,000,000 tokens.