DeepSeek V4 Pro Routes to V4.1 Flash — Same Model, New Price
$0.435/$0.87 input/output (V4 Pro, flat rate)
$0.15/$0.30 input, $0.60/$1.20 output (routed to V4.1 Flash)
What changed
Starting September 14, 2026, all API requests to deepseek-v4-pro will be automatically routed to V4.1 Flash and billed at V4.1 Flash rates. This effectively retires V4 Pro as a separate pricing tier until V4.1 Pro launches.
Before / After
| Before (V4 Pro) | After (routed to V4.1 Flash) | |
|---|---|---|
| Input (cache miss) | $0.435 | $0.15 (off-peak) / $0.30 (peak) |
| Output | $0.87 | $0.60 (off-peak) / $1.20 (peak) |
| Context | 1M | 1M |
This is a 65% price cut for anyone using the V4 Pro endpoint — you get V4.1 Flash quality at V4.1 Flash prices.
What this means for you
If your code calls deepseek-v4-pro: no changes needed. DeepSeek handles the routing transparently. Your bills will drop significantly.
If you were paying $0.435/1M for V4 Pro reasoning: the same endpoint now costs $0.15 off-peak. This is temporary until V4.1 Pro launches.
If you need V4 Pro's specific reasoning capabilities: test whether V4.1 Flash meets your needs. DeepSeek says V4.1 Flash "comprehensively surpassed V4 Pro in performance, cost, speed, and total time."
The full DeepSeek story (25 days)
- Aug 16: Peak/off-peak introduced with 1.57× price increase
- Sep 10: V4.1 Flash launched at $0.15 — cheaper than original flat rate
- Sep 14: V4 Pro routed to V4.1 Flash — 65% price cut for Pro users
Related
- DeepSeek pricing — current rates
- DeepSeek vs ChatGPT