MOSS-Image Models by MOSI AI
MOSI AI releases · 9 in the last 12 months · this family litChangelog →
2 models2026
Last refreshed 2026-10-22. Next refresh: weekly.
Capabilities
VisionAll models
MultimodalAll models
About
OpenMOSS / MOSI AI's open 5B diffusion-transformer text-to-image models with strong Chinese/English text rendering and JSON layout control (Base and SFT).
Decision facts
- Best fit
- image generationvision and multimodal workstructured outputs
- Capability starting point
- MOSS-Image Base with multimodal inputs
- Lowest tracked input
- Not tracked
- Closest related family
- MOSS-Audio
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
2 in view
MOSS-Image BaseCurrent
Use when the workload needs image generation, 5B parameters, and multimodal inputs.
2026-10image generation5B parametersmultimodal inputs
MOSS-Image SFTCurrent
Use when the workload needs image generation, 5B parameters, and multimodal inputs.
2026-10image generation5B parametersmultimodal inputs
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| MOSS-Image Base | Use when the workload needs image generation, 5B parameters, and multimodal inputs. | 2026-10 | image generation5B parametersmultimodal inputs | Current |
| MOSS-Image SFT | Use when the workload needs image generation, 5B parameters, and multimodal inputs. | 2026-10 | image generation5B parametersmultimodal inputs | Current |
Release Timeline
1 release group2026-10
2 current
MOSS-Image Base
Currentimage generation5B parametersmultimodal inputs
MOSS-Image SFT
Currentimage generation5B parametersmultimodal inputs
Specifications(2 models)
| Model | Released | Parameters | Vision | Multimodal |
|---|---|---|---|---|
| MOSS-Image Base | 2026-10 | 5B | Yes | Yes |
| MOSS-Image SFT | 2026-10 | 5B | Yes | Yes |