DeepSeek V3.2 (chat) token cost & pricing
DeepSeek V3.2 is an open-weight general model priced at $0.28 per million input and $0.42 per million output — a ratio so flat that output barely costs more than input. For generation-heavy workloads it is the cheapest credible option in this table.
DeepSeek V3.2 (chat) specifications
| Vendor | DeepSeek |
|---|---|
| Input price | $0.28 / 1M tokens |
| Output price | $0.42 / 1M tokens |
| Cached input price | $0.028 / 1M tokens (90% cheaper) |
| Context window | 128,000 tokens |
| Max output | 8,192 tokens |
| Open weights | Yes |
| Price rank | 2 of 17 (cheapest first) |
| Best for | cheapest, open-weights |
What DeepSeek V3.2 (chat) costs in practice
Four common workloads, priced at list rates. Monthly figures assume 30 days.
| Workload | Per request | Requests/day | Per month |
|---|---|---|---|
| Chatbot · 1K in / 500 out | $0.0005 | 10,000 | $147.00 |
| RAG Q&A · 8K in / 400 out | $0.0024 | 5,000 | $361.20 |
| Doc summary · 30K in / 1K out | $0.0088 | 1,000 | $264.60 |
| Code agent · 60K in / 3K out | $0.02 | 500 | $270.90 |
How many requests does $100 buy?
Assuming a 1,000-token prompt and a 500-token response per request.
How DeepSeek V3.2 (chat) pricing actually works
The near-1:1 input-to-output ratio is the defining feature. Every other vendor here charges 2x to 8x more for output, which means a workload producing long answers can be five to ten times cheaper on DeepSeek than on a Western frontier model. Cached input is $0.028 per million.
When to use DeepSeek V3.2 (chat)
Best for content generation, translation, bulk rewriting, and any pipeline where you are paying mostly for output tokens. Weaker on very hard reasoning and on some English-language nuance; validate on your own evaluation set before committing.
How to cut your DeepSeek V3.2 (chat) bill
Little to optimize on price — the model is already near the floor. Focus instead on whether quality is sufficient: a 10% retry rate on a cheap model can erase its advantage, so measure end-to-end cost per successful output, not per call.
Other DeepSeek models
Cheaper alternatives to DeepSeek V3.2 (chat)
Run the numbers on your own prompt
Paste your actual text into the token counter to get an exact count instead of an estimate. Nothing is uploaded.
Open the calculator →