Ternary-Bonsai-27B-mlx-2bit vs DeepSeek-R1-0528-Qwen3-8B-MLX-4bit
Specs, VRAM requirements and download trends — updated 29 August 2026.
prism-ml · A · Open weights
Params
2.6 B
Context
—
30d
1.0 M
apache-2.0 · commercial OK
lmstudio-community · B · Open weights
Params
1.3 B
Context
131,072
30d
218 K
mit · commercial OK
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | Ternary-Bonsai-27B-mlx-2bit | DeepSeek-R1-0528-Qwen3-8B-MLX-4bit |
|---|---|---|
| Parameters | 2.6 B | 1.3 B |
| Architecture | Qwen3_5ForConditionalGeneration | Qwen3ForCausalLM |
| Context length | — | 131,072 |
| Licence | apache-2.0 | mit |
| Commercial use | Allowed | Allowed |
| Languages | — | — |
| Downloads 30d | 1,012,597 | 218,172 |
| Downloads all time | 4.7 M | 6.0 M |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2026-07-04 | 2025-05-29 |
Download trend
Daily snapshots, last 90 days (13 Aug – 10 Nov)
2.1 M184 K
Ternary-Bonsai-27B-mlx-2bit
DeepSeek-R1-0528-Qwen3-8B-MLX-4bit
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
DeepSeek-R1-0528-Qwen3-8B-MLX-4bit · fp16 @ 128K ctx
8.6 GB / 24 GB
Ternary-Bonsai-27B-mlx-2bit · fp16
10.2 GB / 24 GB