Qwen3.6-35B-A3B Claude 4.7 Opus Reasoning Distilled - MLX oQ4 MTP
MLX/oMLX 4-bit conversion of
r3lax/Qwen3.6-35B-A3B-Claude-4.7-Opus-Reasoning-Distilled
with Qwen MTP tensors preserved and runtime-tested in oMLX.
This is not a new training run or fine-tune. Weights were only converted/quantized for local MLX/oMLX inference.
Quick Facts
Architecture: Qwen3.6 35B-A3B MoE, roughly 3B active parameters per token.
Quantization: oQ4-style MLX 4-bit, group size 64.
MTP: preserved and verified in oMLX.
Test hardware: Apple Silicon M5 Pro with 48GB unified memory.
Runtime tested with oMLX native MTP enabled.
MTP Verification
1 MTP config: mtp_num_hidden_layers=1, mtp_use_dedicated_embeddings=false
2 MTP tensor entries: 42
3 MTP fusion projection: language_model.mtp.fc.weight present and full precision
4 Runtime: oMLX logged native MTP path activation during smoke tests
MTP support is runtime-specific. These tensors are preserved, but non-oMLX runtimes may ignore them.
Local Speed Smoke Test
Measured on an M5 Pro Mac with 48GB unified memory using oMLX.
1 MTP on average: ~89.9 tok/s
2 MTP off average: ~86.0 tok/s
3 Speed lift: ~+4.6% with native MTP enabled
Per-prompt MTP-on smoke results:
1 count120: 160 tokens, 1.726s, 92.69 tok/s, MTP accept 79/79 and 79/79
2 rain180: 220 tokens, 2.465s, 89.26 tok/s, MTP accept 96/123 and 98/121
3 jsoncities: 128 tokens, 1.457s, 87.86 tok/s, MTP accept 64/65 and 64/64
These are local smoke numbers, not a universal benchmark. Prompt, cache state, batching, oMLX version, and hardware will change results.
Usage Notes
Place the model under your oMLX model directory and enable native MTP:
~/.omlx/models/Qwen3.6-35B-A3B-Claude-4.7-Opus-Reasoning-Distilled-MLX-oQ4-MTPRecommended oMLX settings:
1 {
2 "mtp_enabled" : true ,
3 "dflash_enabled" : false ,
4 "turboquant_kv_enabled" : false
5 }
Attribution
Credit to the upstream work:
Qwen/Qwen3.6-35B-A3Blordx64/Qwen3.6-35B-A3B-Claude-4.7-Opus-Reasoning-Distilledr3lax/Qwen3.6-35B-A3B-Claude-4.7-Opus-Reasoning-DistilledAnthropic Claude Opus 4.7 as the teacher model used by the upstream distillation.
oMLX / MLX community for the Apple Silicon runtime and MTP support.
Please credit the upstream authors if you use or redistribute this conversion.
Caveats
MTP runtime activation was verified in oMLX only.
This conversion is experimental.
Reasoning models can emit long <think> traces; set max_tokens intentionally.
The upstream model card notes that distillation transfers reasoning style, not new factual knowledge.
License
Apache-2.0, following the upstream model metadata. Check upstream model cards and any teacher-data usage policies before redistribution or commercial deployment.
