Model benchmarks
qwen2.5:7b-instruct-q4_K_M local LLM performance
As of November 2026, qwen2.5:7b-instruct-q4_K_M runs at up to 20.1 tok/s for local inference (best of 3 community benchmark runs across 1 GPU).
Model size
7.6B
Peak speed
20.1 tok/s
Average speed
19.6 tok/s
Avg PP
3995.0 tok/s
Min memory
4.8 GB
Max context
128,000 tokens
Avg output / run
4,117 tokens
Avg runtime / run
3m 39s
Avg quality
43.4
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 | 39.6 | 43.4 | 48.6 |
| Agent Workflow | 25.5 | 46.0 | 64.1 |
| Code Generation | 26.9 | 33.5 | 38.4 |
| Role Play & Narrative | 50.0 | 51.9 | 54.9 |
| Research & Analysis | 40.1 | 42.1 | 45.1 |
Performance by hardware and tool
Every hardware/tool/quantization combination qwen2.5:7b-instruct-q4_K_M 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 |
|---|---|---|---|---|---|---|---|---|
| Intel Graphics | Ollama | Q4_K_M | 20.1 tok/s | 19.6 tok/s | 4.8 GB | 128,000 tokens | 43.4 | 3 |
Benchmark runs
All 3 qwen2.5:7b-instruct-q4_K_M runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is qwen2.5:7b-instruct-q4_K_M good for coding?
- In our benchmarks, qwen2.5:7b-instruct-q4_K_M scores 33.5/100 for coding. It runs at about 19.6 tok/s, so if you want more speed, tensorfold-qwen38-27b is faster (~25.0 tok/s) and still scores well for coding (84.3/100).
- Is qwen2.5:7b-instruct-q4_K_M good for agentic (tool-using) tasks?
- In our benchmarks, qwen2.5:7b-instruct-q4_K_M scores 46.0/100 for agentic workflows. It runs at about 19.6 tok/s, so if you want more speed, muse-glimmer:latest is faster (~33.3 tok/s) and still scores well for agentic workflows (91.6/100).
- How fast is qwen2.5:7b-instruct-q4_K_M for local inference?
- Across 3 community benchmark runs, qwen2.5:7b-instruct-q4_K_M reaches up to 20.1 tok/s and averages 19.6 tok/s, with the fastest results on Intel Graphics.
- How much memory does qwen2.5:7b-instruct-q4_K_M need?
- The leanest observed configuration used about 4.8 GB of memory (quantizations tested: Q4_K_M).
- Which tools have been used to run qwen2.5:7b-instruct-q4_K_M?
- Benchmarks were submitted using Ollama. Results are community-contributed and updated as new runs arrive.