$ diff deepseek-v4 kimi-k3
DeepSeek V4 vs Kimi K3
DeepSeek and Moonshot AI, side by side. Every value below is the one recorded on the model’s own wiki page, with its citation — nothing is estimated.
DeepSeekDeepSeek V4
- context
- 1M
- weights
- open
- $/M in
- —
- $/M out
- —
- context
- 1M
- weights
- open
- $/M in
- $3
- $/M out
- $15
What actually differs
- Context window
- Identical — both accept 1M tokens, so context length is not a reason to pick either.
- Weights
- Both publish weights — DeepSeek V4 under MIT (open-weight), Kimi K3 under Modified MIT (commercial use permitted; weights publicly downloadable). Check the licences rather than assuming they permit the same commercial use.
Full spec
| Attribute | DeepSeek V4 | Kimi K3 |
|---|---|---|
| Developer | DeepSeek | Moonshot AI |
| Released | 2026-07-16 | 2026-07-16 |
| Context window | 1M tokens (384K max output) | 1M tokens |
| Pricing | Peak: 2× off-peak API rates; off-peak rates not disclosed | $0.30/M cache-hit input · $3.00/M cache-miss input · $15.00/M output |
| License | MIT (open-weight) | Modified MIT (commercial use permitted; weights publicly downloadable) |
| Availability | DeepSeek API, Hugging Face (deepseek-ai/DeepSeek-V4-Pro) | Kimi Code, Kimi app, Kimi API; weights on Hugging Face (`moonshot-ai/kimi-k3`, released 2026-07-27) |
Values come from DeepSeek V4 and Kimi K3, where each is cited to its source. This page states no benchmark result and ranks neither model — it puts two published specifications next to each other. Where a lab has not published a figure, the row says so rather than guessing. How these pages are produced.