Get started
Pricing
TNSA bills usage as you go, in Indian Rupees (₹). Text models are priced per 1M tokens of input and output. Images, audio and search agents are priced per action. NGen 4.3 is the current latest family.
USD prices are approximate (₹83.33 = $1); you're billed in INR.
Text & reasoning models
Priced per 1M tokens, charged separately for input and output. NGen 4.3 is the latest family; earlier families remain available at lower rates. Models marked * are currently available in the Platform's Chat only.
| Model | Input / 1M tokens | Output / 1M tokens |
|---|---|---|
| NGen-4.3 Pro | ₹750 | ₹1,800 |
| NGen-4.1 Pro | ₹750 | ₹1,800 |
| NGen-4.1 | ₹500 | ₹1,000 |
| NGen-4 | ₹117 | ₹583 |
| NGen-4.1 Mini | ₹44 | ₹219 |
| NGen-4 Mini | ₹44 | ₹219 |
| NGen-4.1 Lite | ₹18 | ₹146 |
| NGen-4 Lite | ₹18 | ₹146 |
| NGen-4 Flash | ₹15 | ₹88 |
| NGen-4 Blaze | ₹15 | ₹88 |
| NGen 4.5 Vayu * | ₹250 | ₹800 |
| NGen 4.1 Flash * | ₹15 | ₹88 |
| NGen3.9-Pro | ₹33 | ₹50 |
| NGen3.9-Lite | ₹10 | ₹12 |
| NGen3.9-Max-v2-128K | ₹66 | ₹100 |
| NGen3.9-Max-Stable-v3 | ₹83 | ₹125 |
| NGen3.5-Max | ₹50 | ₹75 |
| NGen3.5-Pro | ₹20 | ₹30 |
| NGen3.5-Lite | ₹5 | ₹7 |
| NGen3.1-Pro | ₹42 | ₹58 |
Agents
Some agents bill per token; others bill per action, such as a search or research query.
| Agent | Input / 1M tokens | Output / 1M tokens | Per action |
|---|---|---|---|
| Code-1 | ₹42 | ₹58 | — |
| Code-2 | ₹64 | ₹92 | — |
| MultiAgent-1 | ₹83 | ₹125 | — |
| WebSearch-1 | — | — | ₹16 / query |
| WebSearch-2 | — | — | ₹24 / query |
| Scientist-1 | — | — | ₹50 / query |
| Scientist-2 | — | — | ₹72 / query |
Image generation
Billed per generated image.
| Model | Price / image |
|---|---|
| IGen-1 | ₹250 |
| IGen-1 Fast | ₹110 |
| IGen-1 Nano | ₹83 |
Audio
Billed per minute of audio processed.
| Model | Price |
|---|---|
| AudioTextSpeech-Pro | ₹100 / minute |
| AudioTextSpeech-1 | ₹33 / minute |
| NGen-Realtime-Omni | ₹96 / 10 minutes |
| NGen-Realtime-Audio | ₹64 / 10 minutes |
How billing works
- Usage is metered per request: input and output tokens are counted and priced at the model's rate.
- Image requests are billed per generated image; audio per minute; search and research agents per query.
- Cached input tokens are billed at a reduced rate compared to fresh input.
- Costs accrue against your API key's usage limit — set a limit to cap spend.
- Track per-key usage on the Usage page.
Example cost
A single request to NGen-4 Mini with 1,000 input tokens and 500 output tokens:
text
input 1,000 / 1,000,000 x ₹44 = ₹0.044
output 500 / 1,000,000 x ₹219 = ₹0.110
total = ₹0.154Prices are in Indian Rupees (₹)
Token rates are per 1M tokens; image rates are per image; audio rates are per minute. These match the rates shown on the Models page.