All Models

DeepSeek AI

DeepSeek V3.2
in Europe

DeepSeek V3.2 is an open‑source LLM that combines efficient sparse attention with a scalable reinforcement‑learning framework, delivering strong reasoning abilities and agentic capabilities. It supports conversational interactions, tool‑calling, and structured output while maintaining high performance on benchmarks such as Terminal‑Bench and MathArena. The model is well‑suited for chat assistants, advanced problem solving, and tool‑driven autonomous agents.

deepseek/deepseek-v3.2
GDPR Compliant Hosted in Europe standard Reasoning

164K

Context Window

55

Speed

90

Intelligence

Capabilities

function calling reasoning chat

Best Use Cases

interactive chat assistants
complex reasoning and problem‑solving tasks
agentic tool‑use and structured output generation
academic and competitive exam preparation

Model Information

Provider
DeepSeek AI
Parameters
685B
License
MIT
HuggingFace
View Model Card

Integration

Start with one API call

OpenAI-compatible API. Use the same SDK you already know. Just swap the model parameter.

terminal
curl https://api.supa.works/openai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <YOUR_API_KEY>" \
  -d '{
    "model": "deepseek/deepseek-v3.2",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
Model ID: deepseek/deepseek-v3.2

Why SUPA

Why host on SUPA?

European Data Residency

All data processed and stored exclusively on European servers. Full GDPR compliance.

OpenAI-Compatible API

Use the same SDK and code you already have. Just change the base URL.

No Infrastructure

No GPUs to manage, no scaling to worry about. Just API calls.

No Data Training

We never use your API requests to train models. Your data stays private.

Start using DeepSeek V3.2

Free sandbox tier available. No credit card required.

Get API Access