gpt-oss-20b (high) vs MiMo V2.6 Flash (Reasoning)
Results updated 14 November 2026 · Both in Terminus-2 · How we benchmark
Verdict
Not comparable yet: MiMo V2.6 Flash (Reasoning) is not yet ranked.
- Cost per task
- $0.017 vs No data
- Time per task
- 7.2 min vs No data
- Tokens per attempt
- 67k vs No data
- Image input
- No vs Yes
Not yet ranked. MiMo V2.6 Flash (Reasoning), Terminus-2: no Earth-Bench V2, GeoBenchX or GS-QA2 score yet. 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.
Charts for each: gpt-oss-20b (high), Terminus-2 · MiMo V2.6 Flash (Reasoning), Terminus-2.
Side by side
| Metric | gpt-oss-20b (high), Terminus-2 | MiMo V2.6 Flash (Reasoning), Terminus-2 |
|---|---|---|
| Geospatial Agent Index: Text only | 33 | No data |
| Geospatial Agent Index: Text + vision | No image input | No data |
| Earth-Bench V2 score | 32 | No data |
| GeoBenchX score | 38 | No data |
| GS-QA2 score | 30 | No data |
| Vision agents score | No image input | No data |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 53 | No data |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 37 | No data |
| OGD4All score (standalone; in no index) | 20 | No data |
| GeoNatureAgent score (standalone; in no index) | No data | No data |
| MultiGlobeQA score (standalone; in no index) | 20 | No data |
| GeoAgentEval score (standalone; in no index) | 57 | No data |
| ImpactMesh score (standalone; in no index) | No data | No data |
| Remote sensing and raster analysis score | 33 | No data |
| Change detection score | 28 | No data |
| GIS analysis score | 26 | No data |
| Climate and time series score | 32 | No data |
| Making maps score | 21 | No data |
| Cost per task | $0.017 | No data |
| Time per task (at 100 output tokens/s) | 7.2 min | No data |
| Tokens per attempt | 67k | No data |
| Turns per attempt | 5.2 | No data |
| Input price per 1M tokens | $0.070 | $0.14 |
| Output price per 1M tokens | $0.30 | $0.28 |
| Context window | 128,000 | 1,048,576 |
| Image input | No | Yes |
| Developer | OpenAI | Xiaomi |
Frequently asked questions
Which scores higher on the Geospatial Agent Index, gpt-oss-20b (high), Terminus-2 or MiMo V2.6 Flash (Reasoning), Terminus-2?
Not comparable yet: MiMo V2.6 Flash (Reasoning), Terminus-2 is not yet ranked (no Earth-Bench V2, GeoBenchX or GS-QA2 score yet).
Which is cheaper per task, gpt-oss-20b (high), Terminus-2 or MiMo V2.6 Flash (Reasoning), Terminus-2?
Not known yet: MiMo V2.6 Flash (Reasoning), Terminus-2 has no cost data.
Which writes fewer output tokens per task, gpt-oss-20b (high), Terminus-2 or MiMo V2.6 Flash (Reasoning), Terminus-2?
Not known yet: MiMo V2.6 Flash (Reasoning), Terminus-2 has no output tokens data.
Which has the larger context window, gpt-oss-20b (high), Terminus-2 or MiMo V2.6 Flash (Reasoning), Terminus-2?
MiMo V2.6 Flash (Reasoning), Terminus-2, 1,048,576 tokens to 128,000 tokens.