Model benchmarks

granite-4.2-30b-MLX-8bit local LLM performance

As of September 2026, granite-4.2-30b-MLX-8bit runs at up to 16.1 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).

oMLX8bit
ShareRedditX

Model size

30B

Peak speed

16.1 tok/s

Average speed

16.0 tok/s

Avg PP

608.5 tok/s

Min memory

31.0 GB

Max context

131.072 tokens

Avg output / run

35.800 tokens

Avg runtime / run

38m 17s

Avg quality

72.1

Benchmark runs

2

GPUs tested

1

Quality by task

Average LLM-judged quality (0–100) with the run-to-run spread shown as a P5–P95 band, overall and for each benchmark task, across all 2 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.

TaskP5 (low)AvgP95 (high)
Overall71.172.173.2
Agent Workflow69.073.177.1
Code Generation49.451.553.5
Role Play & Narrative88.589.390.2
Research & Analysis73.474.776.0

Performance by hardware and tool

Every hardware/tool/quantization combination granite-4.2-30b-MLX-8bit has been benchmarked on, ranked by peak token generation speed. Last updated September 2026.

HardwareToolQuantBest tok/sAvg tok/sMemoryContextQualityRuns
Apple M5 MaxoMLX8bit16.1 tok/s16.0 tok/s31.0 GB131.072 tokens72.12

Benchmark runs

All 2 granite-4.2-30b-MLX-8bit runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.

Frequently asked questions

Is granite-4.2-30b-MLX-8bit good for coding?
In our benchmarks, granite-4.2-30b-MLX-8bit scores 51.5/100 for coding. It runs at about 16.0 tok/s, so if you want more speed, IsValorum/Nex-N2.5-mini-APEX-I-MiniPlus-V2.1-GGUF:1 is faster (~25.7 tok/s) and still scores well for coding (85.0/100).
Is granite-4.2-30b-MLX-8bit good for agentic (tool-using) tasks?
In our benchmarks, granite-4.2-30b-MLX-8bit scores 73.1/100 for agentic workflows. It runs at about 16.0 tok/s, so if you want more speed, granite-4.2-8b-Q4_K_L is faster (~71.6 tok/s) and still scores well for agentic workflows (92.9/100).
How fast is granite-4.2-30b-MLX-8bit for local inference?
Across 2 community benchmark runs, granite-4.2-30b-MLX-8bit reaches up to 16.1 tok/s and averages 16.0 tok/s, with the fastest results on Apple M5 Max.
How much memory does granite-4.2-30b-MLX-8bit need?
The leanest observed configuration used about 31.0 GB of memory (quantizations tested: 8bit).
Which tools have been used to run granite-4.2-30b-MLX-8bit?
Benchmarks were submitted using oMLX. Results are community-contributed and updated as new runs arrive.