docs: Tier 2 polish — variant matrix + quant trade-off

README.md

@@ -2,18 +2,15 @@
 library_name: mlx
 base_model: Qwen/Qwen3.5-27B
 tags:
-- mlx
-- mlx-lm
-- rotorquant
-- kv-cache-quantization
-- qwen3.5
-- thinking-model
-- 8bit
-- quantized
+  - mlx
+  - mlx-lm
+  - rotorquant
+  - kv-cache-quantization
+  - qwen3.5
+  - thinking-model
+  - 8bit
+  - quantized
 license: apache-2.0
-pipeline_tag: text-generation
-language:
-- en
 ---
 
 # Qwen3.5-27B-RotorQuant-MLX-8bit
@@ -81,3 +78,39 @@ Qwen3.5-27B generates extended reasoning before responses by default. The combin
 - [Base model](https://huggingface.co/Qwen/Qwen3.5-27B)
 - [KV-cache-only variant](https://huggingface.co/majentik/Qwen3.5-27B-RotorQuant)
 - [RotorQuant GitHub](https://github.com/scrya-com/rotorquant)
+
+## Quant trade-off (MLX lane)
+
+| Bits | Approx size | Use case | Recommendation |
+|---|---|---|---|
+| 2-bit | ~7.3 GB | Aggressive quantization | Very low-RAM Macs |
+| 3-bit | ~10 GB | Lossy but small | Low-RAM Macs |
+| 4-bit | ~12 GB | Balanced default | Recommended for most Macs |
+| 5-bit | ~14 GB | Higher fidelity | Quality-sensitive |
+| 6-bit | ~17 GB | Approaching FP16 quality | High-fidelity |
+| **8-bit** | ~21 GB | Near-lossless reference | **Fidelity-critical work** |
+
+(Current variant — **8bit** — is bolded.)
+
+## Variants in this family
+
+(Showing 16 sibling variants under `majentik/qwen3.5-27b-*`. The current variant — `RotorQuant-MLX-8bit` — is **bolded**.)
+
+| Variant | Runtime | Approx size | Use case |
+|---|---|---|---|
+| [RotorQuant](https://huggingface.co/majentik/qwen3.5-27b-rotorquant) | runtime modifier | n/a | KV-cache root (weight-agnostic) |
+| [RotorQuant-2bit](https://huggingface.co/majentik/qwen3.5-27b-rotorquant-2bit) | transformers | n/a | Standalone 2-bit weights |
+| [RotorQuant-GGUF-IQ4_XS](https://huggingface.co/majentik/qwen3.5-27b-rotorquant-gguf-IQ4_XS) | llama.cpp | ~23 GB | Lossy 4-bit, low-RAM CPU/edge |
+| [RotorQuant-GGUF-Q2_K](https://huggingface.co/majentik/qwen3.5-27b-rotorquant-gguf-Q2_K) | llama.cpp | ~16 GB | Lossy, low-RAM CPU/edge |
+| [RotorQuant-GGUF-Q3_K_M](https://huggingface.co/majentik/qwen3.5-27b-rotorquant-gguf-Q3_K_M) | llama.cpp | ~21 GB | Smaller 3-bit, CPU-friendly |
+| [RotorQuant-GGUF-Q4_K_M](https://huggingface.co/majentik/qwen3.5-27b-rotorquant-gguf-Q4_K_M) | llama.cpp | ~30 GB | Balanced default |
+| [RotorQuant-GGUF-Q5_K_M](https://huggingface.co/majentik/qwen3.5-27b-rotorquant-gguf-Q5_K_M) | llama.cpp | ~36 GB | Higher fidelity, more RAM |
+| [RotorQuant-GGUF-Q8_0](https://huggingface.co/majentik/qwen3.5-27b-rotorquant-gguf-Q8_0) | llama.cpp | ~57 GB | Near-lossless reference |
+| [RotorQuant-MLX-2bit](https://huggingface.co/majentik/qwen3.5-27b-rotorquant-mlx-2bit) | mlx-lm | ~8.6 GB | Apple Silicon, smallest |
+| [RotorQuant-MLX-4bit](https://huggingface.co/majentik/qwen3.5-27b-rotorquant-mlx-4bit) | mlx-lm | ~17 GB | Apple Silicon balanced |
+| **RotorQuant-MLX-8bit** | mlx-lm | ~32 GB | Apple Silicon reference |
+| [TurboQuant](https://huggingface.co/majentik/qwen3.5-27b-turboquant) | runtime modifier | n/a | KV-cache root (weight-agnostic) |
+| [TurboQuant-2bit](https://huggingface.co/majentik/qwen3.5-27b-turboquant-2bit) | transformers | n/a | Standalone 2-bit weights |
+| [TurboQuant-MLX-2bit](https://huggingface.co/majentik/qwen3.5-27b-turboquant-mlx-2bit) | mlx-lm | ~8.6 GB | Apple Silicon, smallest |
+| [TurboQuant-MLX-4bit](https://huggingface.co/majentik/qwen3.5-27b-turboquant-mlx-4bit) | mlx-lm | ~17 GB | Apple Silicon balanced |
+| [TurboQuant-MLX-8bit](https://huggingface.co/majentik/qwen3.5-27b-turboquant-mlx-8bit) | mlx-lm | ~32 GB | Apple Silicon reference |