DeepSeek-V3.2 vs GPT-5.4 Nano
API pricing, benchmark indices and specs — updated 19 October 2026.
deepseek-ai · A · Open weights
Params
685.4 B
Context
163,840
30d
5.4 M
mit · commercial OK
openai · B · API-only
Input / 1M
$0.20
Output / 1M
$1.25
AA index
20.7
proprietary · 400K context
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | DeepSeek-V3.2 | GPT-5.4 Nano |
|---|---|---|
| Deployment | Open weights · also via API | API-only (closed weights) |
| Parameters | 685.4 B | n/a — API-only |
| Architecture | DeepseekV32ForCausalLM | n/a — API-only |
| Context length | 163,840 | 400,000 |
| Licence | mit | Proprietary |
| Commercial use | Allowed | Per API terms |
| Languages | — | n/a — API-only |
| Downloads 30d | 5,435,964 | n/a — API-only |
| Downloads all time | 30.4 M | n/a — API-only |
| Quantizations on the Hub | SAFETENSORS | n/a — API-only |
| Gated | No | n/a — API-only |
| First seen on the Hub | 2025-12-01 | n/a — API-only |
| API input price / 1M tokens | $0.28 | $0.20 |
| API output price / 1M tokens | $0.42 | $1.25 |
| Est. cost per 1M input + 1M output tokens | $0.70 | $1.45 |
| Cached input / 1M tokens | $0.028 | $0.02 |
| AA Intelligence Index | 21.5 | 20.7 |
| AA Coding Index | 44.2 | 56.1 |
| AA Agentic Index | — | 16 |
| Max output tokens | 65,536 | 128,000 |
| API providers | 11 | 2 |
| Tool calling | Yes | Yes |
| Reasoning mode | Yes | Yes |
API cost
Estimated cost of 1M input + 1M output tokens via OpenRouter
DeepSeek-V3.2
$0.70
GPT-5.4 Nano
$1.45
DeepSeek-V3.2 is 2.1× cheaper at list price.
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