DeepSeek V4 Pro
$0.435
Input / MTok
$0.87
Output / MTok
$0.003625
Cache read / MTok
Moving to peak/off-peak billing 2026-08-16; see scheduled_change.
API identifier deepseek-v4-pro · Context 1,000,000 tokens · Max output 384,000
· vendor pricing page
Scheduled change · 0 days away
On 16 August 2026 pricing moves to $0.66 input / $1.98 output per million tokens. DeepSeek moves to peak/off-peak billing at 16:00 UTC. Off-peak (majority of hours) rate shown here; peak hours 01:00-04:00 and 06:00-10:00 UTC bill at $1.32 input / $3.96 output / $0.044 cache-read, double the off-peak rate..
Full price sheet
| Metric | USD / million tokens |
|---|---|
| Input | $0.435 |
| Output | $0.87 |
| Cache read | $0.003625 |
Price history
| Observed | Metric | Price |
|---|---|---|
| 2026-08-07 12:06:21 | Input | $0.435 |
| 2026-08-07 12:06:21 | Output | $0.87 |
| 2026-08-07 12:06:21 | Cache read | $0.003625 |
Current prices are free and always will be. The full series — every observation since we started watching, which nobody can reconstruct after the fact — is the part Pro pays for.
Compare with
Models at a similar price point.
| Model | Input | Output | |
|---|---|---|---|
| Codestral | $0.3 | $0.9 | compare → |
| Llama 3.3 70B Versatile | $0.59 | $0.79 | compare → |
| Gemma 4 31B Instruct (Pearl AI) | $0.28 | $0.86 | compare → |
| Gemma 4 31B IT | $0.39 | $0.97 | compare → |
| GPT-5.6 Luna | $0.2 | $1.2 | compare → |
| MiniMax M3 | $0.3 | $1.2 | compare → |