Compare
DeepSeek V3.1 vs Deepseek V3 1 Terminus
This page is context-first: how much text each model can take in one request. Full specs adds capabilities and limits; the pricing matrix below is only about $/million tokens from hosts that list both models.
Model
Deepseek V3 1 Terminus
Context window
164K
163,840 tokens · ~123K words
Context window · side by side
Bar length is relative to the larger of the two windows (100% = max of this pair). This is not pricing.
Same context window size for both models.
DeepSeek V3.1 and Deepseek V3 1 Terminus have identical context windows (163K tokens). DeepSeek V3.1 is 25% cheaper on input.
Quick verdicts
Short takeaways — validate with your own workloads.
RAG / high-volume retrieval
Use DeepSeek V3.1. Input tokens are 25% cheaper — critical when sending large retrieved contexts.
Full specs
Context, output, capabilities, and dates. Green highlights the favorable value where we compute a winner.
| Spec | DeepSeek V3.1 | Deepseek V3 1 Terminus |
|---|---|---|
| Context window | 163,840 tokens (163K) | 163,840 tokens (163K) |
| Max output tokens | 163,840 tokens (163K) | 163,840 tokens (163K) |
| Speed tier | Balanced | Balanced |
| Vision | No | No |
| Function calling | Yes | Yes |
| Extended thinking | Yes | Yes |
| Prompt caching | Yes | Yes |
| Batch API | No | No |
| Release date | Aug 2025 | N/A |
Pricing matrix
Dollar rates only: hosts that list both models, per 1M tokens. For how much text fits, use the context section above — not this table.
| Provider | DeepSeek V3.1 in | DeepSeek V3.1 out | Deepseek V3 1 Terminus in | Deepseek V3 1 Terminus out |
|---|---|---|---|---|
| Deepinfra | — | — | $0.270/M | $1.00/M |
| Novita | — | — | $0.270/M | $1.00/M |
| Openrouter | $0.200/M | $0.800/M | — | — |
Frequently asked questions
Powered by Mem0
Use a smaller model.
Get better results.
Mem0 gives your AI long-term memory so you stop re-sending context on every call. That means you can use a smaller, faster, cheaper model — and still get better answers.
Example: a multi-turn chat session
80% less to send — works with any model