Update README.md

README.md

@@ -97,10 +97,8 @@ While this card is **3-bit**, teams often publish multiple precisions. Use this
 
 | Variant | Typical Peak RAM | Relative Speed | Typical Behavior | When to choose |
 |---|---:|:---:|---|---|
-| **2-bit** | ~3.5–4.5 GB | 🔥🔥🔥🔥 | Smallest, most lossy | Absolute minimum RAM / smoke tests |
 | **3-bit** *(this repo)* | **~4.4–8.8 GB** | **🔥🔥🔥🔥** | **Direct, concise**, great latency | **Default** on 8–16 GB Macs |
 | **4-bit** | ~6–8 GB | 🔥🔥🔥 | Better detail retention vs 3-bit | If 3-bit misses small details |
-| **5-bit** | ~8–9.5 GB | 🔥🔥☆ | Higher fidelity | Documents/JSON extraction |
 | **6-bit** | ~7.5–12.5 GB | 🔥🔥 | Best quality under quant | Choose if RAM allows |
 | **8-bit** | ~9.5–12+ GB | 🔥🔥 | Largest quantized size / highest fidelity | When you prefer simpler 8-bit workflows |