Model benchmarks
schematron-3b local LLM performance
As of October 2026, schematron-3b runs at up to 47.6 tok/s for local inference (best of 3 community benchmark runs across 1 GPU).
Model size
3B
Peak speed
47.6 tok/s
Average speed
37.8 tok/s
Avg PP
1304.6 tok/s
Min memory
3.4 GB
Max context
32,768 tokens
Avg output / run
4,536 tokens
Avg runtime / run
2m 36s
Avg quality
39.6
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 | 36.6 | 39.6 | 43.5 |
| Agent Workflow | 30.8 | 59.7 | 80.6 |
| Code Generation | 5.6 | 8.3 | 11.6 |
| Role Play & Narrative | 47.7 | 56.6 | 61.9 |
| Research & Analysis | 24.1 | 33.8 | 48.0 |
Performance by hardware and tool
Every hardware/tool/quantization combination schematron-3b 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 |
|---|---|---|---|---|---|---|---|---|
| NVIDIA GeForce RTX 2050 | LM Studio | Q4_K_S | 47.6 tok/s | 37.8 tok/s | 3.4 GB | 32,768 tokens | 39.6 | 3 |
Benchmark runs
All 3 schematron-3b runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is schematron-3b good for coding?
- In our benchmarks, schematron-3b scores 8.3/100 for coding. It runs at about 37.8 tok/s, so if you want more speed, nvidia/Qwen3.8-27B-NVFP4 is faster (~63.3 tok/s) and still scores well for coding (82.7/100).
- Is schematron-3b good for agentic (tool-using) tasks?
- In our benchmarks, schematron-3b scores 59.7/100 for agentic workflows. It runs at about 37.8 tok/s, so if you want more speed, Nex-N2.5-mini-APEX-Mini is faster (~88.2 tok/s) and still scores well for agentic workflows (91.3/100).
- How fast is schematron-3b for local inference?
- Across 3 community benchmark runs, schematron-3b reaches up to 47.6 tok/s and averages 37.8 tok/s, with the fastest results on NVIDIA GeForce RTX 2050.
- How much memory does schematron-3b need?
- The leanest observed configuration used about 3.4 GB of memory (quantizations tested: Q4_K_S).
- Which tools have been used to run schematron-3b?
- Benchmarks were submitted using LM Studio. Results are community-contributed and updated as new runs arrive.