Image Captioning on COCO pt-PT
50.8AccuracyAMALIA-VL
Evaluation Results
| Method | Links | |
|---|---|---|
| AMALIA-VLOpen weights=true, Fully open vision data=true, Fully open LLM=true2026.06 | 50.8 | |
| AMALIA-VL-DPOOpen weights=true, Fully open vision data=true, Fully open LLM=true2026.06 | 50.3 | |
| LLaVA-OV1.5-8BOpen weights=true, Fully open vision data=true2026.06 | 36.5 | |
| Qwen3.5-9BOpen weights=true2026.06 | 33.7 | |
| MiniCPM-V4Open weights=true2026.06 | 32 | |
| Qwen3-VL-8BOpen weights=true2026.06 | 30.4 | |
| InternVL3.5-8BOpen weights=true2026.06 | 29.9 | |
| Gemma-3-12BOpen weights=true2026.06 | 26.7 | |
| GLM-4.6V-FlashOpen weights=true2026.06 | 25.8 | |
| TowerVision-9BOpen weights=true, Fully open vision data=true2026.06 | 24.4 | |
| EuroVLM-9BOpen weights=true, Fully open LLM=true2026.06 | 23.9 | |
| Ministral-3-8BOpen weights=true2026.06 | 22.4 | |
| Salamandra-VL-7BOpen weights=true, Fully open vision data=true, Fully open LLM=true2026.06 | 19.7 | |
| Molmo2-8BOpen weights=true, Fully open vision data=true, Fully open LLM=true2026.06 | 19.5 | |
| Nemotron-Nano-12BOpen weights=true2026.06 | 18.8 | |
| Perception-LM-8BOpen weights=true, Fully open vision data=true, Fully open LLM=true2026.06 | 15.2 |