AX-Clef-27B-MLX-AXQ-6bit

Research development artifact. Base affine6 packing with higher-precision protection, not MXFP6 or uniformly six-bit storage. No quality certificate is issued.

Property Value
Source Cloudflare/clef
Immutable source revision 0b331204bb13fbd2ca93a64df1956f5b55478ce5
Complete storage 26,484,659,432 bytes
Complete measured BPW 7.708893
Manifest budget class 8bit
MTP No MTP payload is advertised.

The precision distribution and exact per-module settings are recorded in axquant_manifest.json, axquant_plan.json, and config.json. Complete BPW includes protected tensors and storage overhead. File size does not establish peak runtime memory requirements. Converted from the pinned original source with existing precision protection floors.

Observed runtime evidence

The exact runtime payload passed the following scoped check on Metal. The receipt binds its runtime input files. This is a functional smoke, not a quality benchmark, MTEB evaluation, coding evaluation, long-context test or MTP acceleration result.

{
  "task": "decision",
  "runtime": "mlx-lm",
  "version": "0.32.2",
  "mlx_version": "0.32.4",
  "cases": [
    {
      "answers": {
        "department": {
          "choice": "technical",
          "confidence": 0.7495,
          "probabilities": {
            "billing": 0.2505,
            "technical": 0.7495
          },
          "type": "choice"
        },
        "outage": {
          "noul": 0.7585,
          "type": "noul"
        },
        "urgency": {
          "confidence": 0.9546,
          "legend": {
            "0": "Can wait",
            "1": "Today"
          },
          "probabilities": {
            "0": 0.0454,
            "1": 0.9546
          },
          "score": 0.9546,
          "type": "score"
        }
      },
      "passed": true
    }
  ]
}
Runtime Evidence for this artifact
MLX-LM 0.32.2 Passed the task-specific smoke above
MLX-VLM Not separately tested
AX Engine Not tested; architecture metadata is not execution evidence
omlx unsupported (static only)
mtplx unsupported (static only)

Unexercised modalities and MTP behavior remain unverified. Peer runtime-specific exports require distinct variants and their own testing. See the manifest-bound static report.

Use and provenance

python -m pip install -U huggingface_hub mlx-vlm
hf download AutomatosX/AX-Clef-27B-MLX-AXQ-6bit --local-dir ./AX-Clef-27B-MLX-AXQ-6bit

Pin the immutable Hub revision. Use the included clef_mlx.load and model.systemone API for decision heads. The source license and notices are retained where supplied. The upstream repository and its license govern downstream use; no additional commercial grant or upstream benchmark/certification transfer is implied.

Downloads last month
-
Safetensors
Model size
27B params
Tensor type
BF16
·
U32
·
MLX
Hardware compatibility
Log In to add your hardware

6-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for AutomatosX/AX-Clef-27B-MLX-AXQ-6bit

Base model

Qwen/Qwen3.8-27B
Finetuned
Cloudflare/clef
Quantized
(41)
this model

Collections including AutomatosX/AX-Clef-27B-MLX-AXQ-6bit