whatmodel.sh / image / Show-o (show-o-demo-512)
Show-o (show-o-demo-512)
↗ share
+ contribute
Show-o (show-o-demo-512) is an open-weights image model from Show Lab, NUS, released October 2024. Its published size is 1.3B parameters. Licence: apache-2.0.
Model information sources
Release date ↗ · Weights licence ↗ · Parameter count ↗
1.3B = the Show-o transformer (Phi-1.5 1.3B base) as the makers state it; the separate MAGVIT-v2 image tokenizer (showlab/magvitv2) is not included. The 512x512 checkpoint is the same network trained further at higher resolution.
Score within the LMArena text-to-image comparison set only.
$ how it compares
The image models ranked either side of Show-o (show-o-demo-512), out of 197 with a score, then more from Show Lab, NUS.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 157 |  Kandinsky 2.2 | 37±5 | local | ||
| 158 |  Janus-Pro-1B | 36±4 | api $0.0005/image +local | ||
| 159 | Show-o (show-o-demo-512) | 36±3 | local | ||
| 160 |  DALL·E 2 | 36±3 | comparison only | ||
| 161 |  SSD-1B | 36±4 | local | ||
| more from Show Lab, NUS | |||||
| 123 | Show-o2 1.5B | 45±5 | comparison only | ||
| 138 | Show-o2 7B | 42±3 | local | ||
| 146 | Show-o2 1.5B-HQ | 40±3 | local | ||
$ what it's good at
text to image▂▅▅▃▄▄█▃▁▂▁▁ #101 of 126 * · 95% of its score · 1 test
Beats Latent Diffusion text2img-large (LDM-KL-8-G, LAION-400M), VQ-Diffusion (ITHQ), Stable Diffusion XL 0.9 (base) here, though they rank higher overall. Trails PixArt-α (PixArt-XL-2-1024-MS), SSD-1B here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| T2I-ReasonBench (overall) | 42.8% ✎ | ▅▂█▂▂▂····▂▂ beats 2/13 | 95% |
$ where to run it
Other ways to run it
- Run it as published · needs ~7.9 GB, NVIDIA only · from github.com/showlab (official)
git clone https://github.com/showlab/Show-o && cd Show-o && pip3 install -r requirements.txt && wandb login && python3 inference_t2i.py config=configs/showo_demo_512x512.yaml batch_size=1 validation_prompts_file=validation_prompts/showoprompts.txt guidance_scale=5 generation_timesteps=50 mode='t2i'
$ how the score was worked out · T2I-ReasonBench: Benchmarking… 95%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 T2I-ReasonBench: Benchmarking Reasoning-Informed Text-to-Image Generation (Findings of ACL 2026) | 1 | 95% |
| prior · pull toward the middle | 5% | |
| weighted average | = 36.1 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json