Model benchmarks
Spark-X2.6-4B-Precision-42pc local LLM performance
As of October 2026, Spark-X2.6-4B-Precision-42pc runs at up to 62.2 tok/s for local inference (best of 2 community benchmark runs across 2 GPUs).
Model size
4B
Peak speed
62.2 tok/s
Average speed
59.1 tok/s
Avg PP
1110.9 tok/s
Min memory
11.0 GB
Max context
1,048,576 tokens
Avg output / run
132,706 tokens
Avg runtime / run
39m 8s
Avg quality
66.2
Benchmark runs
2
GPUs tested
2
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.
| Task | P5 (low) | Avg | P95 (high) |
|---|---|---|---|
| Overall | 56.9 | 66.2 | 75.4 |
| Agent Workflow | 74.5 | 77.1 | 79.6 |
| Code Generation | 8.9 | 39.6 | 70.3 |
| Role Play & Narrative | 68.9 | 71.1 | 73.4 |
| Research & Analysis | 70.8 | 76.8 | 82.9 |
Performance by hardware and tool
Every hardware/tool/quantization combination Spark-X2.6-4B-Precision-42pc 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 RTX A4000 | llama.cpp | — | 62.2 tok/s | 62.2 tok/s | 61.8 GB | 1,048,576 tokens | 55.9 | 1 |
| NVIDIA GeForce RTX 3060 | llama.cpp | — | 56.0 tok/s | 56.0 tok/s | 11.0 GB | 192,768 tokens | 76.5 | 1 |
Benchmark runs
All 2 Spark-X2.6-4B-Precision-42pc runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is Spark-X2.6-4B-Precision-42pc good for coding?
- In our benchmarks, Spark-X2.6-4B-Precision-42pc scores 39.6/100 for coding. It runs at about 59.1 tok/s, so if you want more speed, Ornith-1.5-35B-A3B.APEX-I-NanoPlus is faster (~110.5 tok/s) and still scores well for coding (81.1/100).
- Is Spark-X2.6-4B-Precision-42pc good for agentic (tool-using) tasks?
- In our benchmarks, Spark-X2.6-4B-Precision-42pc scores 77.1/100 for agentic workflows. It runs at about 59.1 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 Spark-X2.6-4B-Precision-42pc for local inference?
- Across 2 community benchmark runs, Spark-X2.6-4B-Precision-42pc reaches up to 62.2 tok/s and averages 59.1 tok/s, with the fastest results on NVIDIA RTX A4000.
- How much memory does Spark-X2.6-4B-Precision-42pc need?
- The leanest observed configuration used about 11.0 GB of memory.
- Which tools have been used to run Spark-X2.6-4B-Precision-42pc?
- Benchmarks were submitted using llama.cpp. Results are community-contributed and updated as new runs arrive.