Model benchmarks
gemma-4-31b-turbo local LLM performance
As of November 2026, gemma-4-31b-turbo runs at up to 90.5 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
31B
Peak speed
90.5 tok/s
Average speed
88.3 tok/s
Avg PP
2158.6 tok/s
Min memory
n/a
Max context
65,536 tokens
Avg output / run
6,956 tokens
Avg runtime / run
1m 13s
Avg quality
65.4
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 | 63.3 | 65.4 | 67.6 |
| Agent Workflow | 72.0 | 75.5 | 79.0 |
| Code Generation | 52.6 | 54.0 | 55.3 |
| Role Play & Narrative | 68.1 | 68.5 | 68.9 |
| Research & Analysis | 59.6 | 63.8 | 68.0 |
Performance by hardware and tool
Every hardware/tool/quantization combination gemma-4-31b-turbo 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 AI PRO R9700 | radiance | — | 90.5 tok/s | 88.3 tok/s | n/a | 65,536 tokens | 65.4 | 2 |
Benchmark runs
All 2 gemma-4-31b-turbo runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is gemma-4-31b-turbo good for coding?
- In our benchmarks, gemma-4-31b-turbo scores 54.0/100 for coding. It runs at about 88.3 tok/s, so if you want more speed, ornith-1.5-35b-apex is faster (~147.2 tok/s) and still scores well for coding (83.9/100).
- Is gemma-4-31b-turbo good for agentic (tool-using) tasks?
- In our benchmarks, gemma-4-31b-turbo scores 75.5/100 for agentic workflows. It runs at about 88.3 tok/s, so if you want more speed, Ornith-1.5-35B-A3B-BigBang-oQ8e-mtp is faster (~102.7 tok/s) and still scores well for agentic workflows (91.1/100).
- How fast is gemma-4-31b-turbo for local inference?
- Across 2 community benchmark runs, gemma-4-31b-turbo reaches up to 90.5 tok/s and averages 88.3 tok/s, with the fastest results on AMD Radeon AI PRO R9700.
- Which tools have been used to run gemma-4-31b-turbo?
- Benchmarks were submitted using radiance. Results are community-contributed and updated as new runs arrive.