Qwen3.6-35B-A3B-oQ2e-MTP-MLX

Qwen3.6-35B-A3B-oQ2e-MTP-MLX

This is the Qwen3.6-35B-A3B model package used by Sprig in Treeish. It combines an oQ2e mixed-precision MLX quant with its embedded Multi-Token Prediction head and the chat template validated for Sprig's coding-agent workflow.

The package is intended to be downloaded and loaded as-is. Treeish pins an exact repository commit rather than following main.

Model

• Base model: Qwen/Qwen3.6-35B-A3B
• Architecture: 35B total parameters, approximately 3B active parameters
• Quantisation: oQ2e imatrix-enhanced mixed precision
• Default quantisation: 2-bit affine, group size 64
• Per-tensor overrides: 3, 4, 5, 6 and 8-bit
• Format: MLX safetensors
• MTP: one embedded layer under language_model.mtp.*
• Context length: 262,144 tokens

The package contains 2,052 indexed tensors, including 42 embedded MTP tensors. The three weight shards contain 13,563,313,332 bytes of tensor data.

Provenance

The quantised weights and model configuration are byte-identical to jknlsn/Qwen3.6-35B-A3B-oQ2e-mtp at commit 37da6e86c51a2be6fb1259b8b8482e7f3ec41323. That source declares oMLX 0.5.0 as the converter and Qwen/Qwen3.6-35B-A3B as the base model. It does not identify the exact base-model commit used for conversion, so this package should be treated as a curated, byte-pinned distribution rather than a byte-reproducible conversion recipe.

The chat template is Froggeric's Qwen3.6 v21.3 template. It is byte-identical to archive/v21_chat_template.jinja in froggeric/Qwen-Fixed-Chat-Templates at commit 9f14778c92c3b5ed3e0738085694c0d3452802dd, and to the corrected template distributed by mlx-works/Qwen3.6-35B-A3B-oQ2-mtp at commit 3355da833e54bfda1f97dec6b25c411dd9e90084.

No model, tokenizer or configuration tensors were changed for this release. The release adds the corrected chat template, licence, provenance and file manifest.

Runtime compatibility

This package is validated with Treeish's pinned MLX Swift runtime. A different runtime must support the per-tensor quantisation overrides in config.json and the embedded Qwen MTP layout.

Treeish uses the model from 24 GB of unified memory and recommends 32 GB. Actual headroom depends on context length, cache settings and other running applications.

Treeish validation

The release candidate was validated on a 36 GB M4 Max using the release build of Treeish's benchmark:

• All 2,052 indexed tensors were present exactly once and assigned to the expected shard.
• Embedded-MTP tool use produced a parsed search_text call with the requested query and result count.
• The warmed 1,066-token performance fixture generated 86.5 tokens/s without MTP and 110.6 tokens/s with MTP block size 3.
• MTP block size 3 accepted 60 of 69 proposed draft tokens in that fixture.
• Sprig's exact-string edit format produced 8 exact edits from 11 fixtures and 10 structurally valid edits from 11 fixtures in one deterministic-format trial.

These figures describe one machine and one small release fixture. They are not general model benchmarks.

Limitations

This is an aggressive low-bit quantisation. It trades some model quality for memory use and local generation speed. Applications should validate it against their own prompts, tool format and runtime before relying on it.

The model package contains no custom executable code. File sizes, SHA-256 digests and source revisions are recorded in RELEASE_MANIFEST.json.

Licence

Qwen3.6-35B-A3B is licensed under Apache 2.0. The full licence text is included in LICENSE. The Froggeric template repository also declares Apache 2.0 and is attributed above.