DeepSeek V4 Pro vs Kimi K3: price, context and capabilities
DeepSeek V4 Pro (deepseek-v4-pro) and Kimi K3 (kimi-k3) side by side, from XiaoZhan AIhub's live catalogue: prices, context window and capabilities. Both are called with the same API key.
| DeepSeek V4 Pro | Kimi K3 | |
|---|---|---|
| Model ID | deepseek-v4-pro | kimi-k3 |
| Vendor | DeepSeek | Moonshot AI |
| Category | Text generation | Text generation |
| Context window | 128K | 256K |
| Price | Input 0.378Credits/K0.54Credits· | Input 1.2Credits/K· |
| Image input | ||
| Web search | ||
| Tool calling | ||
| Reasoning | ||
| JSON mode | ||
| Open weights | ||
| Protocols | OpenAI Chat Completions | OpenAI Chat Completions · Anthropic Messages |
Key differences
- DeepSeek V4 Pro costs less per input token (about 2.2× cheaper than Kimi K3).
- DeepSeek V4 Pro costs less per output token.
- Kimi K3 has the longer context window (256K vs 128K tokens).
DeepSeek V4 Pro
DeepSeek V4 Pro. 128K context, reasoning_content stream. Billed at DeepSeek's official peak / off-peak prices: full price on weekdays 9:00-12:00 and 14:00-18:00 Beijing time, half price at all other times, including weekends and public holidays.
Prices, examples and API →
Kimi K3
Moonshot Kimi K3. 256K context, with prompt caching.
Prices, examples and API →
FAQ
- Can I call DeepSeek V4 Pro and Kimi K3 with the same API key?
- Yes. Create one API key on XiaoZhan AIhub, set your SDK's base_url to https://api.dflop.top/v1, and switch between them by putting deepseek-v4-pro or kimi-k3 in the model field.
- How are DeepSeek V4 Pro and Kimi K3 billed?
- Each is billed by its own pricing (see the price row above); prices are in credits, 60 credits = ¥1. Failed requests are not charged.