Local Qwen Coding Agent on MacBook: Tackling Context & Output Bottlenecks
Techngro · reddit · 2026-08-08
The author shares severe bottlenecks encountered when using oMLX to run Qwen3.6-35B-A3B-4bit with the Pi coding agent to process a large TypeScript repository on a 32GB MacBook Air.
Core Problems:
- Insufficient Context: System, tool, and repository instruction overhead consumes 10.2K tokens. With a 19.5K context window limit, only 5K tokens of working room are left, frequently triggering auto-compaction and interrupting tasks.
- Output Limit: The 4K max output token limit is too tight for Qwen with thinking enabled, causing incomplete responses.
Optimization Ideas:
- Reduce system instruction overhead and lower post-compaction retained tokens (e.g., decrease keepRecentTokens from 4096 to 1024-2048).
- Disable or reduce thinking mode during tool-heavy repository investigation to save output budget.
- Use different profiles for different tasks (investigation vs. deep coding).
- Explore reliably expanding the effective context window on 32GB Apple Silicon.
More from coding & agent
- Opus 5.5 + Three.js demo stuns with water physics, fish and blossoms — nikola_mr64990 · 2026-10-16
- Grok-built linear regression bot claims $508K PnL in 125 days on Polymarket — Roger_M_Taylor · 2026-10-16
- Karpathy's knowledge engine workflow: Claude Code turns raw notes into a self-updating wiki — nikola_mr64990 · 2026-10-16
- Sierra and Meta unveil Personal Agent Protocol with Zapier as launch partner — sujingshen · 2026-10-16
- ArcKit adds 102 AI-narrated short films for commands across 14 community overlays — mcraddock · 2026-10-16
- Personal agents live or die on chaining tasks, not single-task ability — sujingshen · 2026-10-16