JiRackUltra_1b vs Llama-3.2-3B-Instruct-FP8-dynamic
Specs, VRAM requirements and download trends — updated 24 October 2026.
CMSManhattan · A · Open weights
Params
1.8 B
Context
131,072
30d
891 K
mit · commercial OK
RedHatAI · B · Open weights
Params
3.6 B
Context
131,072
30d
233 K
llama3.2 licence · conditional
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | JiRackUltra_1b | Llama-3.2-3B-Instruct-FP8-dynamic |
|---|---|---|
| Parameters | 1.8 B | 3.6 B |
| Architecture | Qwen2ForCausalLM | LlamaForCausalLM |
| Context length | 131,072 | 131,072 |
| Licence | mit | llama3.2 |
| Commercial use | Allowed | Conditional |
| Languages | 13 | 8 |
| Downloads 30d | 891,277 | 233,446 |
| Downloads all time | 2.6 M | 958 K |
| Quantizations on the Hub | GGUF, SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2026-08-02 | 2024-09-25 |
Download trend
Daily snapshots, last 47 days (25 Sep – 10 Nov)
2.5 M207 K
JiRackUltra_1b
Llama-3.2-3B-Instruct-FP8-dynamic
A overtook B on 26 Sep
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
JiRackUltra_1b · Q4_K_M
2.0 GB / 24 GB
JiRackUltra_1b · fp16 @ 128K ctx
8.7 GB / 24 GB
Llama-3.2-3B-Instruct-FP8-dynamic · fp16 @ 128K ctx
14.0 GB / 24 GB