Flowise with an India AI API (₹ billing)

Use Flowise with ₹-native AI models

Flowise's OpenAI-compatible ChatOpenAI node lets you point to any LLM provider. Swap the base URL and API key to unoblox and run QwenAI, DeepSeek, or Claude at rupee rates on a monthly GST invoice from an Indian entity with input tax credit.

Setup in 2 steps

1. Get your API key: unoblox.ai/sign-in → save ub-gw-…
2. In Flowise, add ChatOpenAI LLM node:
   • Base URL: https://api.unoblox.ai/v1
   • API Key: ub-gw-…
   • Model Name: qwen/qwen3-235b-a22b-instruct-2507 (swap as needed)
   • Test the connection; billing flows to your dashboard

Why Flowise + unoblox works great

• No code needed. Drag nodes, connect them, deploy in Flowise UI
• Open models included. Qwen, Llama, DeepSeek, all priced in ₹
• Monthly invoice. One GST document for accounting; claim input credit
• Privacy. Data stays between Flowise and unoblox gateway (no third vendor)

Model compatibility matrix

Model	₹ input/1M	Vision	Tools	Speed
Qwen3.8-27B	₹16.32/48.96	✓	✓	Fast
DeepSeek V3.2	₹26.21/38.30	–	✓	Deep reasoning
Qwen3 Free	₹0/0	–	✓	Rate-limited
Claude Sonnet	₹201.6/1008	✓	✓	Quality

Common Flowise architectures

• Chatbot: Flowise UI + Qwen (low cost) or Claude (quality)
• RAG pipeline: Retriever (Flowise) → Qwen embeddings (₹0) → Qwen LLM (₹16.32/M)
• Batch processing: Scheduled Flowise job → DeepSeek → webhook callback

FAQ

Q: Does Flowise run on my machine or Flowise Cloud? Both. Self-hosted or Flowise Cloud—your choice. Unoblox endpoint works either way.

Q: Can I use environment variables for the API key? Yes. Set OPENAI_API_KEY=ub-gw-… in Flowise .env and leave the node field blank.

Q: How do I add system prompts in Flowise? In the ChatOpenAI node, set System Prompt field to your instruction (e.g., "You are a tax accountant"). Flowise forwards it as the system role.

Q: What happens if I hit the rate limit? Unoblox returns a 429 with a retry-after header. Flowise's built-in retry logic (exponential backoff) handles it automatically.

Q: Can I switch models mid-conversation? Yes. Create two ChatOpenAI nodes (one per model) and use Flowise conditions to branch (e.g., if task = "reasoning", use DeepSeek).

Q: Is Flowise's memory compatible with unoblox? Fully. Flowise manages conversation history; unoblox just processes each message. No special setup needed.

Get started in rupees → https://unoblox.ai/sign-in