DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window. It is designed for advanced reasoning, coding,...
At a glance
- Context window
- 1Mtokens
- Max output
- 393Ktokens
- Knowledge to
- —
- Released
- 2026
Capabilities
Benchmarks
- Intelligence index
- 31
- Coding index
- 59
- Agentic index
- 28
Design Arena
| Category | Elo | Win rate |
|---|---|---|
| models · 3d | 1,282 | 56.8 |
| models · codecategories | 1,256 | 51.9 |
| models · gamedev | 1,252 | 52.3 |
| models · website | 1,247 | 50.5 |
| models · uicomponent | 1,237 | 50.6 |
| models · dataviz | 1,219 | 48.7 |
| models · asciiart | 1,167 | 46.4 |
| models · svg | 1,166 | 45.4 |
Standard evaluations
| Benchmark | Score | Tasks |
|---|---|---|
| gpqa_diamond | 0.86 | 3,561 |
| tau_bench_verified_airline | 0.76 | 1,048 |
Source: Artificial Analysis, Design Arena · as of Sep 16, 2026
Performance
Uptime on IranRouter
0.6%
Reference uptime
100%
Median latency
1,518 ms
Throughput
40.8 tok/s
IranRouter uptime comes from our own automated probes; the reference figures are daily measurements of this model's upstream routes. Each card shows a 30-day average, and a day without data stays empty.
Popularity
- Current rank
- #18
- Rank change, 7 days
- +4
Daily tokens · 90 days
329.7B
Ranked by daily token volume across the global model market; the figure above the chart is the daily average.
Supported parameters
Parameters this model accepts; anything else is ignored.
Quick start
from openai import OpenAI
client = OpenAI(
api_key="ir-...",
base_url="https://iranrouter.com/v1",
)
resp = client.chat.completions.create(
model="deepseek/deepseek-v4-pro",
messages=[{"role": "user", "content": "سلام!"}],
)
print(resp.choices[0].message.content)Only base_url and the key change; the rest of your code stays as it is.