DeepSeekOpen weights
DeepSeek V4 Flash.
DeepSeek V4 Flash is DeepSeek's mixture-of-experts model with 284 billion parameters, of which 13 billion are active. It suits code, reasoning and agent workflows; the vision variant also understands images, documents and charts with the same text performance.
base_url="https://api.noirdoc.de/v1" model="deepseek/deepseek-v4-flash-max"
- Data centres in Germany
- § 203 confidentiality
- Zero retention
- List prices
Details.
Where the model runs, which rules apply and what it can do.
- Data centres in Germany weber.cloud Contracting party: weber.digital GmbH · based in Balingen, Germany
- § 203 German Criminal Code (StGB)
- Confidentiality agreed with weber.cloud
- Weights
- Open weights
- Features
- Tools, Vision, Reasoning and JSON
Providers.
We offer DeepSeek V4 Flash through these providers.
Provider § 203 model=
€ · 1M tokens Input Cache Read Cache Write Output
weber.cloud DE Own data centres in Germany
deepseek/deepseek-v4-flash-max Max Input 0.50 Cache Read 0.15 Output 1.80
weber.cloud DE Own data centres in Germany
deepseek/deepseek-v4-flash-think Think Input 0.30 Cache Read 0.09 Output 1.20
Get started.
Test selected models right away with €5 starter credit — every model from the first top-up. If you have questions first, talk to our sales team.
import os
from openai import OpenAI
client = OpenAI(
# one line, that's all
base_url="https://api.noirdoc.de/v1",
api_key=os.environ["NOIRDOC_API_KEY"],
)
response = client.chat.completions.create(
model="deepseek/deepseek-v4-flash-max", # provider & model
messages=[{"role": "user", "content": "Hello"}],
)