Nebius AI Studio

Nebius AI Studio is an enterprise AI cloud platform providing ultra-fast, cost-effective inference on high-end NVIDIA H100/H200 infrastructure for leading open-source models.

Website: https://studio.nebius.ai/

---

Getting an API Key

1. Log in to Nebius AI Studio.
2. Navigate to API Keys in your project dashboard.
3. Create and copy your API key (nba-...). Store it securely.

---

Supported Models

Nebius AI Studio provides high-throughput OpenAI-compatible endpoints for top open-weights models:

Deep Reasoning & Planning

• deepseek-ai/DeepSeek-R1 - (Recommended for Planning) Frontier 671B reasoning model for multi-step logic and complex code architecture
• deepseek-ai/DeepSeek-R1-Distill-Llama-70B - Fast reasoning model distilled on Llama 70B
• deepseek-ai/DeepSeek-R1-Distill-Qwen-32B - Distilled mathematical and coding reasoning

Coding & Agentic Execution

• Qwen/Qwen2.5-Coder-32B-Instruct - (Recommended for Coding) Surgical diffs, multi-file refactoring, and tool use
• deepseek-ai/DeepSeek-V3 - 671B MoE flagship foundation model for coding and agent workflows
• mistralai/Mistral-Large-2411 - Mistral's flagship high-capability model

Foundation & General Purpose

• meta-llama/Meta-Llama-3.3-70B-Instruct - Flagship 70B open model for general reasoning and planning
• meta-llama/Meta-Llama-3.1-405B-Instruct - Massive 405B foundation model for high-depth code audits
• meta-llama/Meta-Llama-3.1-8B-Instruct - Ultra-fast lightweight model for quick edits

---

Configuration in PostQode

1. Open PostQode Settings ().
2. Select Nebius AI Studio in the API Provider dropdown.
3. Enter your Nebius API Key (nba-...) from studio.nebius.ai.
4. Select your model from the dropdown or specify the model identifier (e.g. Qwen/Qwen2.5-Coder-32B-Instruct or deepseek-ai/DeepSeek-R1).
5. Click Save Settings to connect.