DeepSeek-V3.2 vs Kimi-K2-Instruct
Specs, VRAM requirements and download trends — updated 10 November 2026.
deepseek-ai · A · Open weights
Params
685.4 B
Context
163,840
30d
7.6 M
mit · commercial OK
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | DeepSeek-V3.2 | Kimi-K2-Instruct |
|---|---|---|
| Deployment | Open weights · also via API | Open weights · also via API |
| Parameters | 685.4 B | 1,026.5 B |
| Architecture | DeepseekV32ForCausalLM | DeepseekV3ForCausalLM |
| Context length | 163,840 | 131,072 |
| Licence | mit | other |
| Commercial use | Allowed | Unknown |
| Languages | — | — |
| Downloads 30d | 7,616,440 | 277,334 |
| Downloads all time | 35.7 M | 4.7 M |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2025-12-01 | 2025-07-11 |
| API input price / 1M tokens | $0.26 | $0.57 |
| API output price / 1M tokens | $0.38 | $2.30 |
| Est. cost per 1M input + 1M output tokens | $0.64 | $2.87 |
| Cached input / 1M tokens | $0.13 | — |
| AA Intelligence Index | — | — |
| AA Coding Index | 44.2 | — |
| AA Agentic Index | — | — |
| Max output tokens | 147,456 | 98,304 |
| API providers | 7 | 1 |
| Tool calling | Yes | Yes |
| Reasoning mode | Yes | No |
API cost
Estimated cost of 1M input + 1M output tokens via OpenRouter
DeepSeek-V3.2
$0.64
Kimi-K2-Instruct
$2.87
DeepSeek-V3.2 is 4.5× cheaper at list price.
Download trend
Daily snapshots, last 90 days (13 Aug – 10 Nov)
7.6 M162 K
DeepSeek-V3.2
Kimi-K2-Instruct
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
DeepSeek-V3.2 · fp16 @ 160K ctx
2,815.1 GB / 24 GB
Kimi-K2-Instruct · fp16 @ 128K ctx
3,596.1 GB / 24 GB