DeepSeek V4 pricing: Flash and Pro
DeepSeek V4 is priced in two time bands: peak, and half price off-peak. This page lists both models' prices, the peak schedule and what a request costs. For the provider overview see the DeepSeek API pricing guide; the wider section is AI costs and pricing.
Prices last verified against official provider pages on . How we verify prices. Written by the Chelure Enterprises Editorial Team.
Prices per million tokens
| Model | Input | Output | Cached input (read) | Cache write (5 min) | Batch API | Announced change |
|---|---|---|---|---|---|---|
| DeepSeek V4.1 Flash | $0.30 | $1.20 | $0.006 | - | - | - |
| DeepSeek V4 Pro | $1.32 | $3.96 | $0.044 | - | - | - |
Verified November 13, 2026 against the provider's pricing page. Compare every provider in the price comparison, or model your own workload in the cost calculator.
What each model is for
| Model | Positioning (provider) | Context window | Max output |
|---|---|---|---|
| DeepSeek V4.1 Flash (deepseek-flash) | Non-thinking and thinking modes; vision supported; concurrency limit 2,500 | 1M tokens | Up to 384K tokens |
| DeepSeek V4 Pro | Non-thinking and thinking modes; vision not supported; concurrency limit 500 | 1M tokens | Up to 384K tokens |
What a request costs
Costs per single request, from the prices above. Prices shown are peak; off-peak is half. Agent step uses cache hits at the peak cache-hit rate.
| Model | Chat turn (1,500 input / 400 output tokens, no caching) | RAG answer (8,000 input / 500 output tokens, no caching) | Agent step (30,000 input / 2,000 output tokens, 80% of input read from cache) | 1,000 chat turns |
|---|---|---|---|---|
| DeepSeek V4.1 Flash | $0.0009 | $0.0030 | $0.0043 | $0.93 |
| DeepSeek V4 Pro | $0.0036 | $0.013 | $0.017 | $3.56 |
Peak and off-peak, both models
| Per million tokens | V4.1 Flash peak | V4.1 Flash off-peak | V4 Pro peak | V4 Pro off-peak |
|---|---|---|---|---|
| Cache hit | $0.006 | $0.003 | $0.044 | $0.022 |
| Cache miss (input) | $0.30 | $0.15 | $1.32 | $0.66 |
| Output | $1.20 | $0.60 | $3.96 | $1.98 |
The main table above uses peak prices, so a job run entirely off-peak costs half of what it shows. Peak hours are 01:00-04:00 and 06:00-10:00 UTC on weekdays, excluding Chinese public holidays; everything else is off-peak. Scheduling batch work outside those windows is the simplest saving, and DeepSeek's cache is automatic (see our prompt caching guide).
Flash at $0.30 input is about 4.4 times cheaper than Pro at $1.32, and it also supports vision, while Pro does not; the choice is Pro's capability against Flash's price and image input. Both have 1M-token context.
Frequently asked questions
How much does DeepSeek V4 cost?
At peak prices, DeepSeek V4.1 Flash is $0.30 per million cache-miss input tokens and $1.20 per million output tokens; V4 Pro is $1.32 and $3.96. Off-peak prices are half.
When are DeepSeek's peak hours?
01:00-04:00 UTC and 06:00-10:00 UTC, Monday to Friday, excluding Chinese public holidays. All other hours, including weekends and Chinese public holidays, are off-peak, at half the price, per DeepSeek's pricing page. All times are UTC.
How cheap are cache hits?
At peak, V4.1 Flash cache hits are $0.006 per million tokens and V4 Pro $0.044. Off-peak they are $0.003 and $0.022.
What is the context length?
1M tokens for both models, with a maximum output of up to 384K tokens.
Does DeepSeek support images?
DeepSeek's pricing page lists vision as supported on V4.1 Flash and not supported on V4 Pro. Use deepseek-flash as the model name; the older names deepseek-v4-flash and deepseek-v4-flash-vision-exp still work and route to V4.1 Flash.
Sources
Primary sources opened on November 13, 2026. Prices come from our verified price table (see the methodology).