Hosted AI model APIs with clear INR pricing

Model API families for developers

Start with a specific model family when your application has a defined coding, reasoning, agent, or generation workload.

• Qwen API — coding, agents, multilingual apps, and streaming.
• DeepSeek API — reasoning and coding model access.
• Gemma API — open model chat completions and streaming.
• GLM API — long-context coding and reasoning workflows.
• OpenAI-compatible API India — one API shape across hosted model families.

Hosted AI model API catalog and pricing

These 27 public API model IDs are present in the current Lumino catalog snapshot. Compare their context limits and INR pricing before integrating; availability and prices can change.

Create one API key for the catalog

Change the base URL, not your application

const client = new OpenAI({ apiKey: process.env.LUMINO_API_KEY, baseURL: "https://api.luminoai.co.in/api/v1" });

Call POST /api/v1/chat/completions with model IDs from the published catalog. Streaming chat completions use the same API key.

Prompt caching with visible usage

Supported models report cached input tokens so repeated system prompts, document context, and agent instructions can be measured. In one repeated-prefix test, the first request used 3,066 prompt tokens with 0 cached and cost 51 paise; the repeated request reported 3,040 cached tokens and cost 12 paise.

Production usage

Lumino has served more than 45 million tokens across production model API workloads.