DeepSeek V4 Flash (high), Terminus-2 vs. Claude Sonnet 5.5 (low), Claude Code
DeepSeek V4 Flash (high), Terminus-2 against Claude Sonnet 5.5 (low), Claude Code: Geospatial Agent Index, each evaluation, cost, time and tokens, on the same tasks and limits. The harnesses differ, so part of any gap can come from the harness. Methodology.
Side by side
| Metric | DeepSeek V4 Flash (high), Terminus-2 | Claude Sonnet 5.5 (low), Claude Code |
|---|---|---|
| Geospatial Agent Index: Text only | 72 | 69 |
| Geospatial Agent Index: Text + vision | No image input | 71 |
| Earth-Bench V2 score | 72 | 56 |
| GeoBenchX score | 65 | 77 |
| GS-QA2 score | 79 | 76 |
| Vision agents score | No image input | 78 |
| GeoAgentBench score (Guided workflows; not in the Agent Index) | 94 | 94 |
| GeoAnalystBench score (Guided workflows; not in the Agent Index) | 89 | 95 |
| OGD4All score (standalone; in no index) | No data | 96 |
| GeoNatureAgent score (standalone; in no index) | No data | 100 |
| MultiGlobeQA score (standalone; in no index) | No data | 77 |
| GeoAgentEval score (standalone; in no index) | No data | 100 |
| Remote sensing and raster analysis score | 78 | 69 |
| GIS analysis score | 77 | 75 |
| Climate and time series score | 73 | 59 |
| Spatial statistics score | 38 | 38 |
| Making maps score | 75 | 84 |
| Recognising infeasible requests score | 58 | 80 |
| Cost per task | $0.065 | $0.043 |
| Time per task (at 100 output tokens/s) | 5.9 min | 0.4 min |
| Tokens per attempt | 341k | 70k |
| Turns per attempt | 15.6 | 3.9 |
| Input price per 1M tokens | $0.44 | $2.00 |
| Output price per 1M tokens | $1.32 | $10.00 |
| Context window | 1,048,576 | 1,000,000 |
| Image input | No | Yes |
| Developer | DeepSeek | Anthropic |
Charts
Geospatial Agent Index
Equal average of Earth-Bench V2, GeoBenchX and GS-QA2, 0 to 100 · Higher is better
Data table: Geospatial Agent Index
| Model | Developer | Geospatial Agent Index |
|---|---|---|
| DeepSeek V4 Flash (high), Terminus-2 | DeepSeek | 72 |
| Claude Sonnet 5.5 (low), Claude Code | Anthropic | 69 |
Cost per task
Average cost per task (USD) · Lower is better
Data table: Cost per task
| Model | Developer | Cost per task |
|---|---|---|
| Claude Sonnet 5.5 (low), Claude Code | Anthropic | $0.043 |
| DeepSeek V4 Flash (high), Terminus-2 | DeepSeek | $0.065 |
Output tokens per task
Average output tokens per attempt, reasoning included · Lower is better
Data table: Output tokens per task
| Model | Developer | Output tokens per task |
|---|---|---|
| Claude Sonnet 5.5 (low), Claude Code | Anthropic | 1k |
| DeepSeek V4 Flash (high), Terminus-2 | DeepSeek | 31k |
Frequently asked questions
Which scores higher on the Geospatial Agent Index, DeepSeek V4 Flash (high), Terminus-2 or Claude Sonnet 5.5 (low), Claude Code?
DeepSeek V4 Flash (high), Terminus-2, 72 to 69.
Which is cheaper per task, DeepSeek V4 Flash (high), Terminus-2 or Claude Sonnet 5.5 (low), Claude Code?
Claude Sonnet 5.5 (low), Claude Code, $0.043 to $0.065.
Which writes fewer output tokens per task, DeepSeek V4 Flash (high), Terminus-2 or Claude Sonnet 5.5 (low), Claude Code?
Claude Sonnet 5.5 (low), Claude Code, 1k to 31k.
Which has the larger context window, DeepSeek V4 Flash (high), Terminus-2 or Claude Sonnet 5.5 (low), Claude Code?
DeepSeek V4 Flash (high), Terminus-2, 1,048,576 tokens to 1,000,000 tokens.
More: DeepSeek V4 Flash (high), Terminus-2 · Claude Sonnet 5.5 (low), Claude Code · All comparisons