DeepSeek-V3.2 vs Hermes-3-Llama-3.1-405B
Specs, VRAM requirements and download trends — updated 25 October 2026.
deepseek-ai · A · Open weights
Params
685.4 B
Context
163,840
30d
5.6 M
mit · commercial OK
NousResearch · B · Open weights
Params
405.9 B
Context
131,072
30d
1 K
llama3 licence · conditional
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | DeepSeek-V3.2 | Hermes-3-Llama-3.1-405B |
|---|---|---|
| Deployment | Open weights · also via API | Open weights · also via API |
| Parameters | 685.4 B | 405.9 B |
| Architecture | DeepseekV32ForCausalLM | LlamaForCausalLM |
| Context length | 163,840 | 131,072 |
| Licence | mit | llama3 |
| Commercial use | Allowed | Conditional |
| Languages | — | 1 |
| Downloads 30d | 5,571,243 | 1,327 |
| Downloads all time | 30.7 M | 958 K |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2025-12-01 | 2024-08-13 |
| API input price / 1M tokens | $0.28 | $1.00 |
| API output price / 1M tokens | $0.42 | $1.00 |
| Est. cost per 1M input + 1M output tokens | $0.70 | $2.00 |
| Cached input / 1M tokens | $0.028 | — |
| AA Intelligence Index | 21.5 | — |
| AA Coding Index | 44.2 | — |
| AA Agentic Index | — | — |
| Max output tokens | 65,536 | 16,384 |
| API providers | 11 | 1 |
| Tool calling | Yes | No |
| Reasoning mode | Yes | No |
API cost
Estimated cost of 1M input + 1M output tokens via OpenRouter
DeepSeek-V3.2
$0.70
Hermes-3-Llama-3.1-405B
$2.00
DeepSeek-V3.2 is 2.9× cheaper at list price.
Download trend
Daily snapshots, last 90 days (28 Jul – 25 Oct)
5.6 M1 K
DeepSeek-V3.2
Hermes-3-Llama-3.1-405B
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
Hermes-3-Llama-3.1-405B · fp16 @ 128K ctx
1,867.4 GB / 24 GB
DeepSeek-V3.2 · fp16 @ 160K ctx
2,815.1 GB / 24 GB