Model benchmarks
unsloth-ud-iq4_xs local LLM performance
As of November 2026, unsloth-ud-iq4_xs runs at up to 81.5 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
Unknown
Peak speed
81.5 tok/s
Average speed
74.0 tok/s
Avg PP
490.4 tok/s
Min memory
n/a
Max context
512,000 tokens
Avg output / run
62,974 tokens
Avg runtime / run
14m 21s
Avg quality
85.1
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.
| Task | P5 (low) | Avg | P95 (high) |
|---|---|---|---|
| Overall | 81.8 | 85.1 | 88.4 |
| Agent Workflow | 91.7 | 92.0 | 92.4 |
| Code Generation | 57.2 | 67.5 | 77.9 |
| Role Play & Narrative | 94.2 | 94.8 | 95.3 |
| Research & Analysis | 84.1 | 86.1 | 88.2 |
Performance by hardware and tool
Every hardware/tool/quantization combination unsloth-ud-iq4_xs has been benchmarked on, ranked by peak token generation speed. Last updated November 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| NVIDIA GeForce RTX 4090 | Strata | — | 81.5 tok/s | 74.0 tok/s | n/a | 512,000 tokens | 85.1 | 2 |
Benchmark runs
All 2 unsloth-ud-iq4_xs runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is unsloth-ud-iq4_xs good for coding?
- In our benchmarks, unsloth-ud-iq4_xs scores 67.5/100 for coding. It runs at about 74.0 tok/s, so if you want more speed, turbo is faster (~124.1 tok/s) and still scores well for coding (85.3/100).
- Is unsloth-ud-iq4_xs good for agentic (tool-using) tasks?
- In our benchmarks, unsloth-ud-iq4_xs scores 92.0/100 for agentic workflows. It runs at about 74.0 tok/s, so if you want more speed, Qwen3.8-27B-IQ5R-V4-DFlash2-TP2-i8KV-r21 is faster (~133.3 tok/s) and still scores well for agentic workflows (91.8/100).
- How fast is unsloth-ud-iq4_xs for local inference?
- Across 2 community benchmark runs, unsloth-ud-iq4_xs reaches up to 81.5 tok/s and averages 74.0 tok/s, with the fastest results on NVIDIA GeForce RTX 4090.
- Which tools have been used to run unsloth-ud-iq4_xs?
- Benchmarks were submitted using Strata. Results are community-contributed and updated as new runs arrive.