DeepSeek V4.1 Flash: early EU access on GreenPT

Bring DeepSeek V4.1 Flash into your workflow while keeping inference in the EU. GreenPT is among the first EU providers to add the model, following DeepSeek’s official release on 10 September 2026. You can reach it through the GreenPT API as deepseek-v4.1-flash, using the same key and endpoint as your other models.

For European teams, early access should be practical: a model you can call from your existing tools, clear euro pricing, and a hosting route you can understand.

A million tokens of context, with reasoning you control

V4.1 Flash combines text and image understanding with a 1M-token context window. It gives you room to work with long documents and code, or ask questions about a screenshot alongside written instructions. Read DeepSeek’s model announcement.

On GreenPT, choose from five reasoning settings: none, minimal, low, medium, and high. Use none for a direct answer. Give a harder task more room to reason and a larger output budget. Thinking consumes that budget too, so a very small max_tokens value can leave no room for the visible answer.

Early access, European hosting

We serve this model on EU infrastructure. Your application connects to GreenPT’s OpenAI-compatible endpoint, under GreenPT’s privacy commitments, including no training on your data. See our processing locations and privacy policy for the details.

That makes it straightforward to try a model shortly after release without adding another provider integration to your application. Change the model ID, run it against representative tasks, and compare the results with the models you use already.

What it costs on GreenPT

Input	Output	Cached input
€0.22	€1.10	€0.011

All prices are per million tokens. Cached input is billed at the lower rate only when the provider reports a cache hit. Repeated prefixes can qualify automatically; caching does not discount output tokens. See prompt caching for how usage is counted.

V4.1 Flash is a separate choice from deepseek-v4-flash-0731, with separate pricing. Adding it does not change your existing model selection or defaults.

Energy estimates you can inspect

The API reports estimated energy and carbon impact per request. V4.1 Flash starts with the same energy configuration as DeepSeek V4 Flash 0731. These are inherited estimates, not a separate measurement of V4.1 Flash. The model guide publishes the coefficients and explains the method.

Make your first call

Set GREENPT_API_KEY to your GreenPT API key, then run:

curl https://api.greenpt.ai/v1/chat/completions \
  -H "Authorization: Bearer $GREENPT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-v4.1-flash",
    "messages": [{ "role": "user", "content": "Explain how a database index speeds up a query." }],
    "reasoning_effort": "none",
    "max_tokens": 512
  }'

Choose a task you know well and compare the result. Create a GreenPT account, explore the model catalogue, or follow the DeepSeek V4.1 Flash guide for reasoning settings and integration details.