Qwen2.5-3B-Instruct vs Qwen3-4B-Thinking-2507-FP8
Specs, VRAM requirements and download trends — updated 17 August 2026.
Qwen · B · Open weights
Params
4.4 B
Context
262,144
30d
174 K
apache-2.0 · commercial OK
Specification comparison
Differences are highlighted; identical values are muted.
| Specification | Qwen2.5-3B-Instruct | Qwen3-4B-Thinking-2507-FP8 |
|---|---|---|
| Parameters | 3.1 B | 4.4 B |
| Architecture | Qwen2ForCausalLM | Qwen3ForCausalLM |
| Context length | 32,768 | 262,144 |
| Licence | other | apache-2.0 |
| Commercial use | Unknown | Allowed |
| Languages | 1 | — |
| Downloads 30d | 2,587,484 | 173,699 |
| Downloads all time | 125.6 M | 2.4 M |
| Quantizations on the Hub | SAFETENSORS | SAFETENSORS |
| Gated | No | No |
| First seen on the Hub | 2024-09-17 | 2025-08-06 |
Download trend
Daily snapshots, last 90 days (13 Aug – 10 Nov)
7.6 M174 K
Qwen2.5-3B-Instruct
Qwen3-4B-Thinking-2507-FP8
VRAM side by side
On RTX 4090 · 24 GB · 8K context unless noted
Qwen2.5-3B-Instruct · fp16
7.8 GB / 24 GB
Qwen3-4B-Thinking-2507-FP8 · fp16 @ 256K ctx
27.4 GB / 24 GB