DeepSeek V3.1open
DeepSeek
56.6#96 of 427 overall
- Input / output
- US$ 0,58 / US$ 1,68 US$ per 1M tokens
- Context
- — tokens
- Release
- 08/21/2025 release date
- Weights
- Open
- Speed
- 0 tok/s first token in 0.00 s
Family history
The 4 versions of this family, from first to newest, by release date. Click a point to open that version.
See the data as a table
| Model | Released | Score | Rank | Elo Arena |
|---|---|---|---|---|
| DeepSeek-V2.5 | 09/06/2024 | — | — | 1294 |
| DeepSeek V3 0324 | 03/25/2025 | 33.6 | #314 | 1375 |
| DeepSeek V3.1 | 08/21/2025 | 56.6 | #96 | 1420 |
| DeepSeek V3.2 | 12/01/2025 | 53.3 | #120 | 1425 |
Scores by source
Scores by sourcepercentile = % of ranked models it beats| Metric | Value | Percentile | Source |
|---|---|---|---|
| Arena text (Elo) weight 2 | 1420 | 57% | LMArena |
| Humanity's Last Exam weight 1 | 14.3 | 62% | Artificial Analysis |
| GPQA Diamond | 77.9 | 64% | Artificial Analysis |
| Long-context reasoning | 56.7 | 48% | Artificial Analysis |
By category
| Category | Score | Rank | Metrics |
|---|---|---|---|
| Reasoning | 58.8 | #157 | Humanity's Last Exam, GPQA Diamond |
| Research | 49.1 | #182 | Long-context reasoning |
| Chat | 53.6 | #84 | Arena text (Elo) |
History
RankingIA score on each collection day (hover to see the rank).
Daily collection started on 10/30/2026. The chart appears from the second day the model is ranked.