Model benchmarks
mtplx-flash-next-bare-speed local LLM performance
As of September 2026, mtplx-flash-next-bare-speed runs at up to 58.1 tok/s for local inference (best of 3 community benchmark runs across 1 GPU).
Model size
Unknown
Peak speed
58.1 tok/s
Average speed
57.5 tok/s
Min memory
n/a
Max context
128,000 tokens
Avg output / run
96,115 tokens
Avg runtime / run
27m 51s
Avg quality
85.0
Benchmark runs
3
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 3 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 | 84.2 | 85.0 | 85.6 |
| Agent Workflow | 81.4 | 84.6 | 86.6 |
| Code Generation | 72.8 | 75.6 | 78.1 |
| Role Play & Narrative | 92.7 | 93.3 | 93.9 |
| Research & Analysis | 84.8 | 86.3 | 87.3 |
Performance by hardware and tool
Every hardware/tool/quantization combination mtplx-flash-next-bare-speed has been benchmarked on, ranked by peak token generation speed. Last updated September 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| Apple M5 Max | vLLM | — | 58.1 tok/s | 57.5 tok/s | n/a | 128,000 tokens | 85.0 | 3 |
Frequently asked questions
- Is mtplx-flash-next-bare-speed good for coding?
- In our benchmarks, mtplx-flash-next-bare-speed scores 75.6/100 for coding. It runs at about 57.5 tok/s, so if you want more speed, mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF is faster (~190.9 tok/s) and still scores well for coding (73.4/100).
- Is mtplx-flash-next-bare-speed good for agentic (tool-using) tasks?
- In our benchmarks, mtplx-flash-next-bare-speed scores 84.6/100 for agentic workflows. It runs at about 57.5 tok/s, so if you want more speed, mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF is faster (~190.9 tok/s) and still scores well for agentic workflows (85.7/100).
- How fast is mtplx-flash-next-bare-speed for local inference?
- Across 3 community benchmark runs, mtplx-flash-next-bare-speed reaches up to 58.1 tok/s and averages 57.5 tok/s, with the fastest results on Apple M5 Max.
- How much memory does mtplx-flash-next-bare-speed need?
- The leanest observed configuration used about n/a of memory.
- Which tools have been used to run mtplx-flash-next-bare-speed?
- Benchmarks were submitted using vLLM. Results are community-contributed and updated as new runs arrive.