gpt-oss-20b vs DeepSeek-R1-Distill-Llama-70B
Specs, VRAM requirements and download trends — updated 6 August 2026.
deepseek-ai · B
Params
70.6 B
Context
131,072
30d
512 K
mit · commercial OK
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | gpt-oss-20b | DeepSeek-R1-Distill-Llama-70B |
|---|---|---|
| Parameters | 21.5 B | 70.6 B |
| Architecture | GptOssForCausalLM | LlamaForCausalLM |
| Context length | 131,072 | 131,072 |
| Licence | apache-2.0 | mit |
| Commercial use | Allowed | Allowed |
| Languages | — | — |
| Downloads 30d | 8,616,555 | 512,372 |
| Downloads all time | 87.8 M | 5.8 M |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2025-08-04 | 2025-01-20 |
Download trend
Daily snapshots, last 10 days (28 Jul – 6 Aug)
8.6 M512 K
gpt-oss-20b
DeepSeek-R1-Distill-Llama-70B
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
gpt-oss-20b · fp16 @ 128K ctx
82.4 GB / 24 GB
DeepSeek-R1-Distill-Llama-70B · fp16 @ 128K ctx
325.0 GB / 24 GB