Taplio - Build a growth routine on LinkedIn
Mini toolsSWE-bench Multimodal
Variant of SWE-bench where issues include screenshots, diagrams and other visual context. Tests multimodal software-engineering ability.
Coding
Multimodal
accuracy
Max 100.0%
Released Oct 2024
1
Results
1
Models scored
70.2%
Top: Deepseek 3.2
70.2%
Median
Best results
Top primary scores; one row per model.
Frontier over time
Each dot is one model result; the line traces the running best score.
Not enough data to plot a trend yet.
All results
Showing one canonical row per model. Show all configurations
| # | Model | Score | Conditions | Eval date | Source | Flags |
|---|---|---|---|---|---|---|
| 1 | Deepseek 3.2 | 70.2% | — | Dec 1, 2025 | paper | primary verified |
