Baidu · 8B · Dense
8-step distilled ERNIE-Image — dense text, posters and multi-panel layouts on 24GB cards Check if your GPU or Mac can run ERNIE-Image Turbo locally — 14.9 GB min, 20.5 GB recommended.
2026-044K context
| Quant | Bits | VRAM | Quality | Status |
|---|---|---|---|---|
| Q2_K | 2 | 14.9 GB | low | — |
| Q3_K_M | 3 | 15.9 GB | moderate | — |
| Q4_K_M | 4 | 16.4 GB | good | — |
| Q5_K_M | 5 | 17.4 GB | good | — |
| Q6_K | 6 | 18.4 GB | excellent | — |
| Q8_0 | 8 | 20.5 GB | excellent | — |
| F16 | 16 | 28.7 GB | lossless | — |
Can I run ERNIE-Image Turbo locally?
- Can I run ERNIE-Image Turbo locally?
- ERNIE-Image Turbo needs about 14.9 GB of memory at a minimum and 20.5 GB recommended. Open this page to grade it against your GPU or Mac, then run it with runai, Ollama or LM Studio.
- How much VRAM does ERNIE-Image Turbo need?
- At Q4_K_M, ERNIE-Image Turbo uses about 16.4 GB of VRAM. Higher quants need more memory; lower quants fit tighter cards with a quality tradeoff.