DeepSeek costs $10.12 per month ✓ Official price checked Oct 6, 2026 · official pricing page. That is a chatbot workload on DeepSeek-V4.1-Flash (800 input tokens, 300 output tokens, 20,000 requests, 40% of input cached).
DeepSeek stores account-holder personal data in the People's Republic of China, and the API terms do not warrant uninterrupted service. Privacy policyOpen Platform terms
Cached input on DeepSeek-V4.1-Flash is $0.006 per 1M tokens instead of $0.30. Official note: Peak-hour price; 50% off outside peak hours.
800 input · 300 output · 20,000 requests/month · 40% of input cached
★ Used for the estimate above
DeepSeek-V4.1-Flash
$10.12/month
✓ Official price checked Oct 6, 2026 · Official pricing
Peak-hour price; 50% off outside peak hours.
Good for high-volume chat on a small budget. Not good for work that needs a flagship.
DeepSeek-V4-Pro
$36.71/month
✓ Official price checked Oct 6, 2026 · Official pricing
Peak-hour price; 50% off outside peak hours.
Good for long context at a lower flagship rate. Not good for batch API jobs, because it has no batch discount.
Standard rates. Cached input uses this scenario's cached share. Open the full calculator →
| Model | Provider | Tier | Checked | ||||||
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek-V4-Pro DeepSeek · $1.32 in / $3.96 out · Peak-hour price Peak-hour price; 50% off outside peak hours. | $0.0022 | DeepSeek | Flagship | $1.32 | $3.96 | $0.044 | 1M | Oct 6, 2026 | $0.0022 |
| DeepSeek-V4.1-Flash DeepSeek · $0.30 in / $1.20 out · Peak-hour price Peak-hour price; 50% off outside peak hours. | $0.0006 | DeepSeek | Small | $0.30 | $1.20 | $0.006 | 1M | Oct 6, 2026 | $0.0006 |
Standard pay-as-you-go rates. * Higher rates apply above a context-length threshold. Click a date to open the official pricing page it was checked against. Your cost is one request at the usage selected above.
Other models that cost less than DeepSeek-V4.1-Flash for the default chatbot workload (800 input tokens, 300 output tokens, 20,000 requests). The DeepSeek estimate above can change when you pick another scenario; this list stays on that default.
| Alternative | Monthly cost | Savings | Good for |
|---|---|---|---|
| Qwen3.7-Flash Alibaba (Qwen) | $1.11 | 89% | — |
| Ministral 3 3B Mistral | $1.62 | 84% | — |
| Ministral 3 8B Mistral | $2.44 | 76% | — |
| Ministral 3 14B Mistral | $3.25 | 68% | — |
DeepSeek-V4.1-Flash costs $10.12 per month for a chatbot workload of 800 input tokens, 300 output tokens, and 20,000 requests, with 40% of input tokens cached. That uses the rates checked on Oct 6, 2026. Across DeepSeek models in this table, the same workload runs from $10.12 to $36.71 per month.
DeepSeek input prices in this table run from $0.30 to $1.32 per 1M tokens, and output prices run from $1.20 to $3.96. Checked Oct 6, 2026 against DeepSeek's official pricing page.
The input and output columns are standard rates. Cached input on DeepSeek-V4.1-Flash is $0.006 per 1M tokens instead of $0.30. Official note: Peak-hour price; 50% off outside peak hours.
DeepSeek-V4.1-Flash is the lowest DeepSeek monthly cost for the default chatbot workload, at $10.12 per month ($0.30 input and $1.20 output per 1M tokens).
The DeepSeek rows were checked on Oct 6, 2026. Each row links to the official pricing page the rate was copied from.
DeepSeek's privacy policy says it directly collects, processes, and stores your personal data in the People's Republic of China. The Open Platform terms say that policy covers personal information you provide as the account holder. Personal information of end users of an application you build on the API is not covered by that policy; the terms say you disclose your own processing rules to those users. Privacy policy: https://cdn.deepseek.com/policies/en-US/deepseek-privacy-policy.html. Open Platform terms: https://cdn.deepseek.com/policies/en-US/deepseek-open-platform-terms-of-service.html.
The Open Platform terms provide the API on an "as is" and "as available" basis. They do not warrant that use will be uninterrupted, timely, secure, or free from error, and they say features in a testing period may be unstable and are not guaranteed. Those terms do not state a numeric rate limit. Open Platform terms: https://cdn.deepseek.com/policies/en-US/deepseek-open-platform-terms-of-service.html.