Model benchmarks
qwen35_122b local LLM performance
As of August 2026, qwen35_122b runs at up to 15.2 tok/s for local inference (best of 5 community benchmark runs across 1 GPU).
LM Studio
Model size
122B
Peak speed
15.2 tok/s
Average speed
15.2 tok/s
Min memory
59.6 GB
Max context
6,674 tokens
Best quality
88.4
Benchmark runs
5
GPUs tested
1
Performance by hardware and tool
Every hardware/tool/quantization combination qwen35_122b has been benchmarked on, ranked by peak token generation speed. Last updated August 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| NVIDIA GeForce RTX 3080 | LM Studio | — | 15.2 tok/s | 15.2 tok/s | 59.6 GB | 6,674 tokens | 66.7 | 5 |
Frequently asked questions
- How fast is qwen35_122b for local inference?
- Across 5 community benchmark runs, qwen35_122b reaches up to 15.2 tok/s and averages 15.2 tok/s, with the fastest results on NVIDIA GeForce RTX 3080.
- How much memory does qwen35_122b need?
- The leanest observed configuration used about 59.6 GB of memory.
- Which tools have been used to run qwen35_122b?
- Benchmarks were submitted using LM Studio. Results are community-contributed and updated as new runs arrive.