DeepSeek, open weights
DeepSeek-V3.1-Terminus
DeepSeek-V3.1-Terminus by DeepSeek ranks 97th of 354 ranked models on the Noometry Index as of October 2026, with a score of 43.1. Its strongest category is multilingual, where it ranks 92nd. API pricing starts at $0.27 per million input tokens and $1 per million output tokens, with a 164K-token context window.
Last verified
Specifications
- Noometry rank
- #97 of 354
- Index score
- 43.1
- Evidence
- Confirmed 16 results
- Provider
DeepSeek- Released
- September 22, 2025
- Weights
- Open weights
- Reasoning
- Yes
- Context window
- 164K
- Max output
- 147K
- Input price
- $0.27 / M
- Output price
- $1 / M
- Blended price
- $0.45 / M
- Output speed
- 30 tokens/s Kagi
- Value
- #64 of 219
- Knowledge cutoff
- Unknown
- Input
- text
- Hugging Face
- deepseek-ai/DeepSeek-V3.1-Terminus
Category scores
Each category score combines every public result we have in that category.
- Coding 42.0
- Reasoning 26.4
- Math 38.5
- Multilingual 52.1
- Instruction Following 74.0
- Long Context 43.4
- Writing & Preference 61.0
| Category | Score | Rank | Results |
|---|---|---|---|
| Coding | 42.0 | #113 | 2 |
| Reasoning | 26.4 | #133 | 5 |
| Math | 38.5 | #137 | 1 |
| Multilingual | 52.1 | #92 | 1 |
| Instruction Following | 74.0 | #106 | 1 |
| Long Context | 43.4 | #97 | 1 |
| Writing & Preference | 61.0 | #92 | 3 |
Strengths and weaknesses
Categories where DeepSeek-V3.1-Terminus places highest and lowest among the models ranked in each, with its score against that category's median.
Strongest categories
| Category | Score | vs median | Rank |
|---|---|---|---|
| Writing & Preference | 61.0 | +7.2 | #92 of 312, top 30% |
| Multilingual | 52.1 | +4.7 | #92 of 297, top 31% |
| Long Context | 43.4 | +2.5 | #97 of 296, top 33% |
Weakest categories
| Category | Score | vs median | Rank |
|---|---|---|---|
| Math | 38.5 | +1.9 | #137 of 327, top 42% |
| Reasoning | 26.4 | +2.8 | #133 of 350, top 38% |
| Instruction Following | 74.0 | +2.8 | #106 of 305, top 35% |
Closest competitors
The models ranked just above and below DeepSeek-V3.1-Terminus. When scores are this close, price and speed are often the better way to choose.
| Model | Rank | Score | Blended $/M | Speed | |
|---|---|---|---|---|---|
| MiMo-V2.5 | #93 | 43.4 | $0.18 | — | Compare |
| Kimi K2.7 Code | #94 | 43.3 | $1.71 | — | Compare |
| Qwen3-VL 235B-A22B | #95 | 43.2 | $1.22 | — | Compare |
| Seed 2.0 Pro | #96 | 43.2 | $1.13 | — | Compare |
| Hunyuan Vision 1.5 | #98 | 43.1 | — | — | Compare |
| Mistral Large 4 | #99 | 43.1 | $1.03 | — | Compare |
| Claude Opus 4 | #100 | 43.1 | $30 | 29 | Compare |
| Amazon Nova Experimental Chat 11 10 | #101 | 43.0 | — | — | Compare |
Sponsored placements are available on pages like this one. Advertise on Noometry
Benchmark results
Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.
Coding
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| SciCode | 40.6% | #75 of 121, top 62% | Epoch AI | ||
| LMArena Coding | 1426 | #115 of 294, top 40% | thinking | LMArena | 2026-10-08 |
| ALE-Bench | 745.17 | #66 of 105, top 63% | Epoch AI |
Reasoning
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| Kagi LLM Benchmark | 57.4% | #47 of 99, top 48% | Kagi LLM Benchmark | ||
| CritPt | 1.7% | #73 of 134, top 55% | Epoch AI | ||
| LMArena Hard Prompts | 1426 | #96 of 297, top 33% | thinking | LMArena | 2026-10-08 |
| DTBench | 81.3% | #69 of 151, top 46% | Epoch AI | ||
| LMCA | 28.6% | #80 of 125, top 64% | Epoch AI |
Math
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Math | 1402 | #124 of 285, top 44% | LMArena | 2026-10-08 |
Multilingual
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Non-English | 1407 | #92 of 297, top 31% | LMArena | 2026-10-08 | |
| LMArena Russian | 1436 | #55 of 283, top 20% | thinking | LMArena | 2026-10-08 |
Instruction Following
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Instruction Following | 1404 | #99 of 298, top 34% | thinking | LMArena | 2026-10-08 |
Long Context
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Longer Query | 1421 | #91 of 291, top 32% | thinking | LMArena | 2026-10-08 |
Writing & Preference
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Text | 1419 | #98 of 297, top 33% | thinking | LMArena | 2026-10-08 |
| LMArena Creative Writing | 1403 | #73 of 295, top 25% | LMArena | 2026-10-08 | |
| LMArena Multi-Turn | 1411 | #109 of 295, top 37% | thinking | LMArena | 2026-10-08 |
API pricing by provider
| Route | Input $/M | Output $/M | Cached input $/M | Checked |
|---|---|---|---|---|
| openrouter | $0.27 | $1 | — | 2026-10-10 |
Compare DeepSeek-V3.1-Terminus
- DeepSeek-V3.1-Terminus vs Seed 2.0 Pro
- DeepSeek-V3.1-Terminus vs Hunyuan Vision 1.5
- DeepSeek-V3.1-Terminus vs Qwen3-VL 235B-A22B
- DeepSeek-V3.1-Terminus vs Mistral Large 4
- DeepSeek-V3.1-Terminus vs Kimi K2.7 Code
- DeepSeek-V3.1-Terminus vs Claude Opus 4
- DeepSeek-V3.1-Terminus vs GPT-6 Astra
- DeepSeek-V3.1-Terminus vs Claude Fable 5.1
- DeepSeek-V3.1-Terminus vs Gemini 3.8 Flash
- DeepSeek-V3.1-Terminus vs Kimi K3
- DeepSeek-V3.1-Terminus vs Grok 4.6
- DeepSeek-V3.1-Terminus vs Qwen3.8 Max
- DeepSeek-V3.1-Terminus vs GLM-5.3
- DeepSeek-V3.1-Terminus vs Muse Spark 1.3
Other DeepSeek models
- DeepSeek V4 Pro54.3
- DeepSeek V4 Flash53.6
- DeepSeek V4.1 Flash52.8
- DeepSeek-V3.2-Exp44.3
- DeepSeek-V3.142.8
- DeepSeek-R142.3
- DeepSeek-V3.2-Speciale39.7
- DeepSeek-V339.5
Frequently asked questions
How good is DeepSeek-V3.1-Terminus?
DeepSeek-V3.1-Terminus by DeepSeek ranks 97th of 354 ranked models on the Noometry Index as of October 2026, with a score of 43.1. Its strongest category is multilingual, where it ranks 92nd. API pricing starts at $0.27 per million input tokens and $1 per million output tokens, with a 164K-token context window.
How much does DeepSeek-V3.1-Terminus cost?
DeepSeek-V3.1-Terminus costs $0.27 per million input tokens and $1 per million output tokens on openrouter.
What is DeepSeek-V3.1-Terminus's context window?
DeepSeek-V3.1-Terminus accepts up to 164K tokens of input and can write up to 147K tokens in one response.
Is DeepSeek-V3.1-Terminus open source?
Yes. DeepSeek-V3.1-Terminus's weights are downloadable from Hugging Face (deepseek-ai/DeepSeek-V3.1-Terminus); check the license for commercial terms.
How fast is DeepSeek-V3.1-Terminus?
DeepSeek-V3.1-Terminus generated about 30 output tokens per second in the Kagi LLM Benchmark's timed runs. Speed varies by provider, load and reasoning effort.
What are DeepSeek-V3.1-Terminus's strengths and weaknesses?
Relative to other ranked models, DeepSeek-V3.1-Terminus places best in writing & preference, multilingual, long context and lowest in math, reasoning, instruction following.
What is DeepSeek-V3.1-Terminus best at?
Its best category is multilingual, where it ranks 92nd on Noometry.