MAI-Image-2.7 vs SenseNova U1.5 TailSteer
Newest: MAI-Image-2.7 · Self-hosting: SenseNova U1.5 TailSteer · Most API hosts: MAI-Image-2.7. Not enough arena votes yet for a quality score, so this is a spec comparison.
Microsoft
MAI-Image-2.7
- Arena1,475
- From—
- Max size—
SenseTime
SenseNova U1.5 TailSteer
- Arena—
- From—
- Max size—
Add a model
Make it a three-way comparison.
Where they differ
Quality is the arena score from blind, side-by-side human votes. Price, length and resolution come from each lab’s own listings.
- NewestMAI-Image-2.7MAI-Image-2.7 Nov 17, 2026 · SenseNova U1.5 TailSteer Nov 4, 2026
- Widest inputsSame inputsMAI-Image-2.7: Text prompts, Reference images · SenseNova U1.5 TailSteer: Text prompts, Reference images
- Image editingBoth edit imagesTakes a reference image as input: MAI-Image-2.7 yes · SenseNova U1.5 TailSteer yes
- Self-hostingSenseNova U1.5 TailSteerDownloadable weights
- Most API hostsMAI-Image-2.7MAI-Image-2.7: 1 · SenseNova U1.5 TailSteer: 0
Every spec in one table
Dashes mean the data is not published.
| Specification | ||
|---|---|---|
| Task | — | Multi-turn image editing |
| Published capabilities | — | Full U1.5 checkpoint tuned for stable editing across successive revisions. |
| Model size | — | 8B model family |
| Access | — | Public weights and the SenseNova U1 editing code |
| Sources | — | Publisher documentation · Model weights |
| Quality (blind human votes) | ||
| Text to image arenaElo-style score · higher is better | 1,387 ±12#4 of 86 · 3,082 votes | — |
| Image editing arenaElo-style score · higher is better | 1,475 ±11#4 of 64 · 3,007 votes | — |
| Output | ||
| Produces | Images | Images |
| Inputs | ||
| Text prompts | Yes | Yes |
| Reference images (editing) | Yes | Yes |
| Access | ||
| Weights | Proprietary | Open |
| API hosts | 1 | — |
| Released | Nov 17, 2026 | Nov 4, 2026 |
Which should you choose?
Which is better, MAI-Image-2.7 or SenseNova U1.5 TailSteer?
Neither model has enough arena votes for a reliable quality score yet, so this page compares specs only. Run the same three or four prompts through each and judge the results.
What is the difference between MAI-Image-2.7 and SenseNova U1.5 TailSteer?
MAI-Image-2.7 is an image model from Microsoft (Nov 17, 2026), taking text prompts and reference images and producing images. SenseNova U1.5 TailSteer is an image model from SenseTime (Nov 4, 2026), taking text prompts and reference images and producing images.
Which is newer?
MAI-Image-2.7 Nov 17, 2026; SenseNova U1.5 TailSteer Nov 4, 2026.
Which can edit an existing image?
Takes a reference image as input: MAI-Image-2.7 yes; SenseNova U1.5 TailSteer yes.
Can I run any of these on my own hardware?
SenseNova U1.5 TailSteer. Downloadable weights.