Model benchmarks
strata_iq4 local LLM performance
As of November 2026, strata_iq4 runs at up to 66.0 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
Unknown
Peak speed
66.0 tok/s
Average speed
60.1 tok/s
Avg PP
250.0 tok/s
Min memory
113.1 GB
Max context
32,768 tokens
Avg output / run
67,472 tokens
Avg runtime / run
18m 25s
Avg quality
83.8
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 | 80.8 | 83.8 | 86.8 |
| Agent Workflow | 92.6 | 92.9 | 93.3 |
| Code Generation | 45.0 | 58.2 | 71.3 |
| Role Play & Narrative | 93.3 | 94.4 | 95.5 |
| Research & Analysis | 89.4 | 89.8 | 90.2 |
Performance by hardware and tool
Every hardware/tool/quantization combination strata_iq4 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 |
|---|---|---|---|---|---|---|---|---|
| AMD Radeon RX 7900 XTX | OpenAI-compatible | — | 66.0 tok/s | 60.1 tok/s | 113.1 GB | 32,768 tokens | 83.8 | 2 |
Benchmark runs
All 2 strata_iq4 runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is strata_iq4 good for coding?
- In our benchmarks, strata_iq4 scores 58.2/100 for coding. It runs at about 60.1 tok/s, so if you want more speed, unsloth/Qwen3.8-27B-GGUF:IQ3_S is faster (~104.3 tok/s) and still scores well for coding (83.0/100).
- Is strata_iq4 good for agentic (tool-using) tasks?
- In our benchmarks, strata_iq4 scores 92.9/100 for agentic workflows. It runs at about 60.1 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 strata_iq4 for local inference?
- Across 2 community benchmark runs, strata_iq4 reaches up to 66.0 tok/s and averages 60.1 tok/s, with the fastest results on AMD Radeon RX 7900 XTX.
- How much memory does strata_iq4 need?
- The leanest observed configuration used about 113.1 GB of memory.
- Which tools have been used to run strata_iq4?
- Benchmarks were submitted using OpenAI-compatible. Results are community-contributed and updated as new runs arrive.