whatmodel.sh / image / Janus-Pro-1B
Janus-Pro-1B
↗ share
+ contribute
Janus-Pro-1B is an open-weights image model from DeepSeek, released January 2025. Its published size is 1.5B parameters. Licence: deepseek.
Model information sources
Release date ↗ · Weights licence ↗ · Parameter count ↗
1.5B = LLM parameters (DeepSeek-LLM 1.5B base) as DeepSeek's Janus-Pro report lists them for Janus-Pro-1B (the name says 1B); the SigLIP-Large understanding encoder, VQ generation tokenizer and adaptors/heads are not included.
Score within the LMArena text-to-image comparison set only.
$ how it compares
The image models ranked either side of Janus-Pro-1B, out of 197 with a score, then more from DeepSeek.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 156 |  SDXL-Lightning (4-step UNet) | 37±3 | local | ||
| 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 | ||
| more from DeepSeek | |||||
| 153 | Janus-Pro-7B | 38±2 | api $0.0020/image +local | ||
| 162 | JanusFlow-1.3B | 36±6 | comparison only | ||
| 176 | Janus-1.3B | 32±4 | local | ||
$ what it's good at
text to image▂▅▅▃▄▄█▃▁▂▁▁ #102 of 126 · 94% of its score · 3 tests
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, Show-o (show-o-demo-512) here, though they rank lower overall.
$ where to run it
Through an API
api · DeepInfra · $0.0005/image · deepinfra.com ↗ · price sourceUSD per generated image on DeepInfra (model page header '$0.0005 / image'; catalogue pricing type image_units at 0.05 cents, not deprecated). Product: DeepInfra inference API, text-to-image. API model: deepseek-ai/Janus-Pro-1B. Settings: default output size, one image. Price basis: published per-image list rate. Currency: USD. Not a DeepSeek direct tariff. DeepInfra official model page, observed 2026-09-26.
Other ways to run it
- Run it as published · needs ~5.5 GB, NVIDIA only · from github.com/deepseek-ai
git clone https://github.com/deepseek-ai/Janus && cd Janus && pip install -e . && sed -i.bak 's#deepseek-ai/Janus-1.3B#deepseek-ai/Janus-Pro-1B#; s#"role": "User"#"role": "<|User|>"#; s#"role": "Assistant"#"role": "<|Assistant|>"#' generation_inference.py && python generation_inference.py
Weights and documentation: huggingface.co/deepseek-ai ↗
$ how the score was worked out · DeepSeek-AI — Janus-Pro techn… 63%, WISE: A World Knowledge-Infor… 30%
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 |
|---|---|---|
 DeepSeek-AI — Janus-Pro technical report · lab-reported | 2 | 63% |
| WISE: A World Knowledge-Informed Semantic Evaluation for Text-to-Image Generation (arXiv v4) | 1 | 30% |
| prior · pull toward the middle | 6% | |
| weighted average | = 36.2 |
$ 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