DeepSeek V4 Flash API
Very low-cost DeepSeek V4 model for fast chat, retrieval, and high-volume long-context tasks.
Provider: DeepSeek | Model ID: deepseek-v4-flash | Context Window: 1M tokens
On this model: 1M tokens context window, chat workload class, routed through DeepSeek infrastructure.
Pricing
₹10/1M input · ₹30/1M output. Pay-as-you-go micro-metered billing. UPI and cards accepted with zero minimum commitment.
| Token Type | Price (INR) |
|---|---|
| Input Tokens | ₹10/1M tokens |
| Output Tokens | ₹30/1M tokens |
| Prompt Caching Input | ₹10/1M tokens |
Why chat volume is predictable to budget
Chat workloads have short, stable output. That makes per-call cost easy to forecast: input tokens times the document size, output tokens times the reply length. The main leak is retries and regenerate buttons, which bill as fresh requests.
OpenAI-Compatible Integration
Drop into existing OpenAI SDK code by changing the base URL to https://api.luminoai.co.in/v1. The request shape below is the chat and general purpose pattern.
curl https://api.luminoai.co.in/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $LUMINO_API_KEY" \
-d '{
"model": "deepseek-v4-flash",
"messages": [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Summarise this document in five bullet points."}
]
}'Good fit for chat and general purpose
- Chat assistants and customer-facing support drafting
- Summarisation, extraction, and classification at volume
- Style and tone rewriting for content pipelines
- Any high-volume task where cost per call is the binding constraint
Model Specifications
| Model Identifier | deepseek-v4-flash |
|---|---|
| Provider | DeepSeek |
| Category | chat |
| Context Window | 1M tokens |
| Output Pricing Tier | budget |
| API Protocol | OpenAI-compatible REST API (JSON) |
| Endpoint | https://api.luminoai.co.in/v1/chat/completions |
DeepSeek V4 Flash API FAQ
How much does DeepSeek V4 Flash cost per call?
Multiply your input tokens by the input rate and your reply length by the output rate. Both rates are in the pricing table above.
Is DeepSeek V4 Flash good enough for production support?
For drafting, summarising, and structured extraction, yes. Keep a human in the loop for anything that commits money or changes account state.
How do I control spend on a chat product?
Cap max_tokens, rate-limit the regenerate button, and cache a stable system prompt.
Explore all models in catalog | API Documentation | Test DeepSeek V4 Flash in Playground