Model benchmarks
Frontis-MA1-35B-Q4_K_M local LLM performance
As of October 2026, Frontis-MA1-35B-Q4_K_M runs at up to 14.6 tok/s for local inference (best of 5 community benchmark runs across 1 GPU).
Model size
35B
Peak speed
14.6 tok/s
Average speed
14.5 tok/s
Avg PP
81.8 tok/s
Min memory
n/a
Max context
262,144 tokens
Avg output / run
15,783 tokens
Avg runtime / run
19m 3s
Avg quality
72.4
Benchmark runs
5
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 5 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.
| Task | P5 (low) | Avg | P95 (high) |
|---|---|---|---|
| Overall | 70.1 | 72.4 | 74.3 |
| Agent Workflow | 67.2 | 74.3 | 78.5 |
| Code Generation | 65.2 | 69.5 | 74.3 |
| Role Play & Narrative | 72.9 | 75.8 | 81.4 |
| Research & Analysis | 59.9 | 69.8 | 77.1 |
Performance by hardware and tool
Every hardware/tool/quantization combination Frontis-MA1-35B-Q4_K_M has been benchmarked on, ranked by peak token generation speed. Last updated October 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| Intel Graphics | ik_llama.cpp:cpu | Q4_K_M | 14.6 tok/s | 14.5 tok/s | n/a | 262,144 tokens | 72.4 | 5 |
Benchmark runs
All 5 Frontis-MA1-35B-Q4_K_M runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is Frontis-MA1-35B-Q4_K_M good for coding?
- In our benchmarks, Frontis-MA1-35B-Q4_K_M scores 69.5/100 for coding. It runs at about 14.5 tok/s, so if you want more speed, byteshape/Qwen3.8-27B-GGUF:IQ4_XS-ASCII-Condensed is faster (~42.0 tok/s) and still scores well for coding (86.3/100).
- Is Frontis-MA1-35B-Q4_K_M good for agentic (tool-using) tasks?
- In our benchmarks, Frontis-MA1-35B-Q4_K_M scores 74.3/100 for agentic workflows. It runs at about 14.5 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 Frontis-MA1-35B-Q4_K_M for local inference?
- Across 5 community benchmark runs, Frontis-MA1-35B-Q4_K_M reaches up to 14.6 tok/s and averages 14.5 tok/s, with the fastest results on Intel Graphics.
- Which tools have been used to run Frontis-MA1-35B-Q4_K_M?
- Benchmarks were submitted using ik_llama.cpp:cpu. Results are community-contributed and updated as new runs arrive.