Model comparison
DeepSeek V3.1 Terminus vs DeepSeek V3
Capability radar, value bars, and side-by-side posture.
DeepSeek V3.1 Terminus
Not scored
value score
DeepSeek V3
87
value score
DeepSeek V3.1 Terminus ctx
128K
DeepSeek V3 ctx
128K
Capability radar
Value · context · multimodal · openness · speed posture
Value head-to-head
- DeepSeek V387
Attribute tape
| Field | DeepSeek V3.1 Terminus | DeepSeek V3 |
|---|---|---|
| Developer | ~deepseek | ~deepseek |
| Context | 128K | 128K |
| Modalities | Text, Code | |
| Openness | Open Weights | Open Weights |
| Speed | — | — |
| Price | $0.27/1M in · $1.1/1M out | $0.27/1M in · $1.1/1M out |
| Value score | — | 87 |
| Summary | DeepSeek V3.1 Terminus is an incremental refinement of DeepSeek's V3.1 large language model, offering improved multilingual stability and enhanced performance on code and search agentic workflows. It supports both thinking and non-thinking inference modes and is available via DeepSeek's API and as open weights on Hugging Face. | Strong open-weights MoE general model. |