Claude Fable 5.1 (high) vs GPT-6 Sol (medium)
Results updated 3 November 2026 · In Claude Code and Codex, so part of any gap can come from the harness · How we benchmark
Verdict
Not comparable yet: GPT-6 Sol (medium) is not yet ranked.
- Cost per task
- $0.44 vs $0.047
- Time per task
- 1.1 min vs 0.3 min
- Tokens per attempt
- 132k vs 86k
- Image input
- No data vs Yes
Claude Fable 5.1 (high), Claude Code scores higher on GeoBenchX and GS-QA2; GPT-6 Sol (medium), Codex scores higher on Earth-Bench V2 and Vision agents. GPT-6 Sol (medium), Codex is cheaper and faster: $0.047 against $0.44 per task, and 0.3 min against 1.1 min.
Not yet ranked. GPT-6 Sol (medium), Codex. Not in the Agent Index, cost and time charts until it has a score on every index evaluation. Results so far are in the side-by-side table.
Where they differ most
By kind of work, share of tasks passed, 0 to 100 · Higher is better
| Kind of work | Claude Fable 5.1 (high) | GPT-6 Sol (medium) |
|---|---|---|
| Making maps | 79 | 100 |
| Change detection | 64 | 75 |
| GIS analysis | 88 | 83 |
| Remote sensing and raster analysis | 87 | 89 |
Where the tasks are set: each kind of work links to its tasks and scoresCharts for each: Claude Fable 5.1 (high), Claude Code · GPT-6 Sol (medium), Codex.
Side by side
| Metric | Claude Fable 5.1 (high), Claude Code | GPT-6 Sol (medium), Codex |
|---|---|---|
| Geospatial Agent Index: Text only | 80 | 80 |
| Geospatial Agent Index: Text + vision | 82 | No data |
| Earth-Bench V2 score | 72 | 77 |
| GeoBenchX score | 80 | 75 |
| GS-QA2 score | 90 | 87 |
| Vision agents score | 88 | 91 |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 100 | 100 |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 95 | No data |
| OGD4All score (standalone; in no index) | 97 | No data |
| GeoNatureAgent score (standalone; in no index) | 100 | No data |
| MultiGlobeQA score (standalone; in no index) | 79 | No data |
| GeoAgentEval score (standalone; in no index) | 100 | No data |
| ImpactMesh score (standalone; in no index) | No data | No data |
| Remote sensing and raster analysis score | 87 | 89 |
| Change detection score | 64 | 75 |
| GIS analysis score | 88 | 83 |
| Climate and time series score | 76 | 77 |
| Making maps score | 79 | 100 |
| Cost per task | $0.44 | $0.047 |
| Time per task (at 100 output tokens/s) | 1.1 min | 0.3 min |
| Tokens per attempt | 132k | 86k |
| Turns per attempt | 5.4 | 6.4 |
| Input price per 1M tokens | $10.00 | $2.00 |
| Output price per 1M tokens | $50.00 | $10.00 |
| Context window | 200,000 | 1,050,000 |
| Developer | Anthropic | OpenAI |