Qwen-Image-2.1 vs Qwen-Image-3.0
In image generation, Qwen-Image-2.1 scores 50.6 and Qwen-Image-3.0 48.1 on the 0–100 quality scale. Qwen-Image-2.1 has no listed API price; Qwen-Image-3.0 costs $0.030 per image.
Qwen-Image-2.1Qwen-Image-3.0
Image generation: quality vs price
Compared models are ringed; the other 36 ranked here are greyed.
Lower price is better. Pareto frontier: Z-Image Turbo, Muse Image, MAI-Image-2.6-Flash, MAI-Image-2.6, GPT Image 2.5 Sunburst. 1 models are not plotted: Qwen-Image-2.1.
1 model with no price data — shown in the strip at the left edge
Qwen-Image-2.1
- Best-value frontier (nothing is both cheaper and better)
- Evidencestrong → weak
- Estimated from other categories
- Hover a dot for details · click a lab to highlight it
- Qwen-Image-2.1
- Quality
- 50.6
- Rank
- #16/38
- Price
- —
- Speed
- —
vs Qwen-Image-3.0: +2.5 quality · price n/a
- Qwen-Image-3.0
- Quality
- 48.1
- Rank
- #18/38
- Price
- $0.030/img
- Speed
- 6.2s
vs Qwen-Image-2.1: −2.5 quality · price n/a
Overview
Lab
Alibaba Qwen
Alibaba Qwen
Released
Sep 20, 2026
Jul 21, 2026
Weights
Open
Proprietary
Price per image
—
$0.030
Time per image
—
6.2s
Quality by category
0–100 within each category (100 = best tracked model). Rank is among all models ranked there. Click a row to focus it.
Image generation benchmarks
Rank among models with a published score, and the leaderboard around each model. ~ italic = no published score, estimated from related benchmarks.
AA Text-to-Image Arena
third-party
~#17/36~1071Estimated
estimated from related benchmarks
AA Image Editing Arena
third-party
LMArena Text-to-Image
third-party
~#14/31~1231Estimated
estimated from related benchmarks
LMArena Image Edit
third-party
#13/241367Verified
arena.ai · 2026-09-21
~#14/25~1339Estimated
estimated from related benchmarks
GenEval
day 0
~#3/4~86.0%Estimated
estimated from related benchmarks
~#3/4~86.0%Estimated
estimated from related benchmarks
DPG-Bench
day 0
~#3/4~87.1%Estimated
estimated from related benchmarks
~#3/4~87.2%Estimated
estimated from related benchmarks