PowerMoE-3b vs Qwen3-30B-A3B-Instruct-2507-AWQ-4bit
Specs, VRAM requirements and download trends — updated 30 July 2026.
ibm-research · A · Open weights
Params
3.4 B
Context
4,096
30d
171 K
apache-2.0 · commercial OK
cyankiwi · B · Open weights
Params
30.0 B
Context
262,144
30d
142 K
apache-2.0 · commercial OK
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | PowerMoE-3b | Qwen3-30B-A3B-Instruct-2507-AWQ-4bit |
|---|---|---|
| Parameters | 3.4 B | 30.0 B |
| Architecture | GraniteMoeForCausalLM | Qwen3MoeForCausalLM |
| Context length | 4,096 | 262,144 |
| Licence | apache-2.0 | apache-2.0 |
| Commercial use | Allowed | Allowed |
| Languages | — | — |
| Downloads 30d | 170,877 | 142,312 |
| Downloads all time | 12.0 M | 3.3 M |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2024-08-14 | 2025-07-29 |
Download trend
Daily snapshots, last 90 days (13 Aug – 10 Nov)
1.6 M142 K
PowerMoE-3b
Qwen3-30B-A3B-Instruct-2507-AWQ-4bit
A overtook B on 28 Oct
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
PowerMoE-3b · fp16
15.9 GB / 24 GB
Qwen3-30B-A3B-Instruct-2507-AWQ-4bit · fp16 @ 256K ctx
164.4 GB / 24 GB