Compare models on your machine
Not generic leaderboards: every model is rated for your hardware, your tool and your workload — so you can see which one will actually feel better to use.
MacBook Pro M4 Pro 48 GB · Agentic coding / medium repository
- Qwen3.6 35B-A3B5-bit · LM StudioComfortableSpeed 50–70 tok/sUses 25 GBFree 12 GB
- Qwen3.5 35B-A3B5-bit · LM StudioComfortableSpeed 50–70 tok/sUses 25 GBFree 12 GB
- GLM-4.7-Flash (30B-A3B)6-bit · LM StudioComfortableSpeed 35–50 tok/sUses 26 GBFree 10 GB
- Gemma 4 12B6-bit · LM StudioAcceptableSpeed 17–25 tok/sUses 11 GBFree 25 GB
| Model | Confidence | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Qwen3.6 35B-A3B5-bit · LM Studio | Comfortable | 25 GB | 12 GB | Excellent50–70 tok/s | Comfortable | 128K | Strong | Strong | Medium confidence |
| Qwen3.5 35B-A3B5-bit · LM Studio | Comfortable | 25 GB | 12 GB | Excellent50–70 tok/s | Comfortable | 128K | Good | Good | Medium confidence |
| GLM-4.7-Flash (30B-A3B)6-bit · LM Studio | Comfortable | 26 GB | 10 GB | Comfortable35–50 tok/s | Comfortable | 128K | Strong | Strong | Medium confidence |
| Gemma 4 12B6-bit · LM Studio | Acceptable | 11 GB | 25 GB | Slow17–25 tok/s | Slow | 128K | Fair | Fair | Medium confidence |
Gemma 4 12B
6-bit · LM Studio
Usable for agentic coding, but expect compromises: generation (17–25 tok/s) is slow for agentic coding — slow steps compound across many agent iterations; ingesting a prompt of 18K tokens takes ≈75 s and each step ≈35 s; across ~25 calls a task takes ≈15 min.
≈25 GB headroom
17–25 tok/s
≈75 s first prompt
32K context
Official
Partly suited
Qwen3.6 35B-A3B
5-bit · LM Studio
Yes — this should be comfortable for agentic coding.
≈12 GB headroom
50–70 tok/s
≈20 s first prompt
32K context
Official
Suited
Qwen3.5 35B-A3B
5-bit · LM Studio
Yes — this should be comfortable for agentic coding.
≈12 GB headroom
50–70 tok/s
≈20 s first prompt
32K context
Official
Suited
GLM-4.7-Flash (30B-A3B)
6-bit · LM Studio
Yes — this should be comfortable for agentic coding.
≈10 GB headroom
35–50 tok/s
≈25 s first prompt
32K context
Official
Suited