Model Comparison

Short guidance for choosing models.

Quick Picks

• Local default: qwen2.5‑coder:7b (balanced for most laptops)
• Higher quality local: qwen3‑coder:30b (requires more memory)
• Cloud: pick based on cost and latency targets

Notes

• Results depend on hardware, model settings, and task type.
• Use local models for privacy and predictable cost.