Llama-3.1-8B-Instruct vs avibe
Specs, VRAM requirements and download trends — updated 6 August 2026.
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | Llama-3.1-8B-Instruct | avibe |
|---|---|---|
| Parameters | 8.0 B | 7.9 B |
| Architecture | LlamaForCausalLM | Qwen3ForCausalLM |
| Context length | — | 32,768 |
| Licence | llama3.1 | apache-2.0 |
| Commercial use | Conditional | Allowed |
| Languages | 8 | 2 |
| Downloads 30d | 8,014,793 | 173,439 |
| Downloads all time | 175.3 M | 208 K |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | Yes — accept licence | No |
| First seen on the Hub | 2024-07-18 | 2025-10-20 |
Download trend
Daily snapshots, last 10 days (28 Jul – 6 Aug)
8.0 M159 K
Llama-3.1-8B-Instruct
avibe
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
avibe · fp16
19.1 GB / 24 GB
Llama-3.1-8B-Instruct · fp16
19.4 GB / 24 GB