jedisct1/Qwen3.6-35B-rust-v2-MTP-8bit.mlx

Qwen3.6-35B-rust-v2 MTP 8-bit MLX

A Rust-focused Qwen3.6-35B-A3B model for Apple Silicon, packaged in MLX with native MTP enabled.

Use it as a coding assistant for Rust projects: generating focused patches, explaining diffs, tightening tests, reading command output, and making small repo-aware edits. It was tested with Swival on local tool-calling workflows.

This is the 8-bit native-MTP variant. It is a good choice when you want more precision than the 4-bit package while keeping quantized memory use.

Which Variant Should I Use?

• Use this repo if your runtime supports native MTP for this Qwen3.5 MoE MLX layout and you want the 8-bit package.
• Use jedisct1/Qwen3.6-35B-rust-v2-MTP-4bit.mlx if you want the smallest and fastest MTP variant.
• Use jedisct1/Qwen3.6-35B-rust-v2-MTP-bf16.mlx if you want the full-precision MTP package and have enough memory.
• Use jedisct1/Qwen3.6-35B-rust-v2-8bit.mlx if your loader does not support native MTP sidecars.

What It Is Good At

• Writing idiomatic Rust patches from a concise change request.
• Explaining Rust diffs in commit-message style.
• Following tool-calling workflows where it needs to inspect files before editing.
• Keeping changes focused instead of turning small fixes into broad rewrites.
• Working with tests, compiler errors, command output, and repository context.

Example Prompts

Generate a patch:

You are an expert Rust developer.

Generate a focused patch for this change:

Replace the unwrap() calls in parse_config() with proper error propagation, and add a regression test for invalid TOML.

Explain a change:

You are an expert Rust developer.

Describe what this Rust diff changes and call out any behavior change:

<paste a Rust diff here>

Use tools:

Inspect the repository, find where request timeout errors are handled, update the Rust code to preserve the original error source, and run the relevant tests.

Compatibility

This repo intentionally keeps native MTP weights. If your loader rejects language_model.mtp.* tensors or mtp.safetensors, use the plain 8-bit sibling instead:

jedisct1/Qwen3.6-35B-rust-v2-8bit.mlx

Limitations

• Outputs should be reviewed before use, especially unsafe code, concurrency code, and changes that affect security boundaries.
• The model works best on focused Rust changes, tests, and explanations. Very large refactors may need to be split into smaller steps.
• Tool calling depends on the runtime and client preserving the chat template and tool schema format.