DeepSeek · released Aug 13, 2026
DeepSeek V4 Pro
DeepSeek V4 Pro is an open-weight model with good coding results and a low price.
Open weights3 months oldThinks before answering
- Input price
- $0.66 / M
- Output price
- $1.98 / M
- Context
- 1M tokens
- Longest reply
- 393K
- Reads
- text
- Knows up to
- –
Where it shines
- It scores 77.6 on SWE-bench.
- It costs $0.66 in, $1.98 out per million tokens.
Where it struggles
- Its web dev score of 1464 trails newer open models.
Where you can use it
- Hugging FaceDownloadWeights are public.
- DeepSeek APIAPI$0.66 in / $1.98 out per million tokens.
- OpenRouterAPIListed as
deepseek/deepseek-v4-pro
How it scores on each job
Scores run 0 to 100 and blend every measure we have for that job. The number after the dot is its rank among the 33 models in the lineup.
The evidence
| Measure | Score | Rank in lineup | Source |
|---|---|---|---|
| Epoch Capabilities Index | 156 | #20 | Epoch AI |
| LMArena text | 1451 | #18 | LMArena |
| LMArena hard prompts | 1461 | #21 | LMArena |
| LMArena instruction following | 1445 | #21 | LMArena |
| LMArena multi-turn | 1468 | #15 | LMArena |
| LMArena creative writing | 1446 | #15 | LMArena |
| LMArena writing & language | 1447 | #18 | LMArena |
| LMArena coding | 1469 | #23 | LMArena |
| LMArena math | 1455 | #23 | LMArena |
| LMArena business & finance | 1443 | #20 | LMArena |
| LMArena legal & government | 1464 | #18 | LMArena |
| LMArena medicine | 1469 | #15 | LMArena |
| WebDev Arena | 1464 | #25 | LMArena |
| SWE-bench Verified | 78% | #2 | Epoch AI |
| GPQA Diamond | 92% | #15 | Epoch AI |
| FrontierMath | 65% | #22 | Epoch AI |
| SimpleQA Verified | 53% | #12 | Epoch AI |
| ARC-AGI-2 | 61% | #18 | Epoch AI |
Compare it with these
NewOpen weights
DeepSeek
DeepSeek V4 Flash
DeepSeek retired V4 Flash on Sep 10, and its API name now points to V4.1 Flash.
- Price
- $0.15 in · $0.60 out
- Context
- 1M
Open weights7 weeks old
DeepSeek
DeepSeek V4.1 Flash
DeepSeek V4.1 Flash is the cheapest strong open model and builds websites well.
Websites & appsCoding
- Price
- $0.045 in · $0.50 out
- Context
- 1M
Open weights3 months old
Z.ai
GLM-5.3
GLM-5.3 is Z.ai's open-weight model and handles chat decently at a low price.
Everyday questions
- Price
- $1.2 in · $4.18 out
- Context
- 1M