Kimi K3 (Reasoning) vs GPT-6 Sol (medium)
Results updated 31 October 2026 · In Kimi 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
- $0.00 vs $0.047
- Time per task
- 1.5 min (estimated) vs 0.3 min
- Tokens per attempt
- 345k vs 86k
- Image input
- Yes vs Yes
GPT-6 Sol (medium), Codex scores higher than Kimi K3 (Reasoning), Kimi Code on all 3 evaluations scored for both. Kimi K3 (Reasoning), Kimi Code is cheaper ($0.00 against $0.047 per task), but GPT-6 Sol (medium), Codex is faster (0.3 min against 1.5 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 | Kimi K3 (Reasoning) | GPT-6 Sol (medium) |
|---|---|---|
| Recognising infeasible requests | 52 | 75 |
| Making maps | 82 | 100 |
| Spatial statistics | 38 | 50 |
| Remote sensing and raster analysis | 81 | 88 |
Where the tasks are set: each kind of work links to its tasks and scoresCharts for each: Kimi K3 (Reasoning), Kimi Code · GPT-6 Sol (medium), Codex.
Side by side
| Metric | Kimi K3 (Reasoning), Kimi Code | GPT-6 Sol (medium), Codex |
|---|---|---|
| Geospatial Agent Index: Text only | 75 | 80 |
| Geospatial Agent Index: Text + vision | No data | No data |
| Earth-Bench V2 score | 74 | 77 |
| GeoBenchX score | 65 | 75 |
| GS-QA2 score | 85 | 87 |
| Vision agents score | No data | 91 |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 96 | 100 |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 100 | 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) | 77 | No data |
| GeoAgentEval score (standalone; in no index) | 96 | No data |
| Remote sensing and raster analysis score | 81 | 88 |
| GIS analysis score | 84 | 83 |
| Climate and time series score | 78 | 76 |
| Spatial statistics score | 38 | 50 |
| Making maps score | 82 | 100 |
| Recognising infeasible requests score | 52 | 75 |
| Cost per task | $0.00 | $0.047 |
| Time per task (at 100 output tokens/s) | 1.5 min (estimated) | 0.3 min |
| Tokens per attempt | 345k | 86k |
| Turns per attempt | No data | 6.4 |
| Input price per 1M tokens | $3.00 | $2.00 |
| Output price per 1M tokens | $15.00 | $10.00 |
| Context window | 1,000,000 | 1,050,000 |
| Image input | Yes | Yes |
| Developer | Moonshot AI | OpenAI |