FreeToken benchmarks
An inference engine for running open-weight LLMs locally.
9 community reports
FreeToken results by GPU
Hardware with reported FreeToken results, with each report's model, quant and speed, newest first. Runs on several cards, with speculative decoding, with batched requests or with part of the model in system RAM say so, since each describes a different setup. Prompt processing is how fast the input is read, not how long you wait for the first word. How to measure time to first token.
FreeToken on NVIDIA RTX 50904 reports
- Qwen3.8 125B · 6B active · NVFP426.2 t/s(part in system RAM)
- Qwen3.8 125B · 6B active · NVFP450.5 t/s(part in system RAM)
- Qwen3.8 125B · 6B active · NVFP450.0 t/s, 2,300 t/s prompt(part in system RAM)
- Qwen3.8 125B · 6B active73.6 t/s
FreeToken on NVIDIA RTX 30902 reports
- DeepSeek V4 Flash 284B · 13B active · fp85.6 t/s(2 cards, part in system RAM)
- Qwen3.8 125B · 6B active · NVFP448.0 t/s(2 cards, part in system RAM)
FreeToken on Intel Arc B580 12GB1 report
- Qwen3.8 27B · UD-Q2_K_XL0.7 t/s
FreeToken on NVIDIA RTX 4060 Laptop 8GB1 report
- Qwen3.6 35B · 3B active39.0 t/s(part in system RAM)
FreeToken, GPU not identified1 report
- Qwen3.820.0 t/s(part in system RAM)
This engine doesn't yet have an editorial profile on llamaperf. The community reports below show how it's been used in practice across different hardware.
Compare GPUs running FreeToken
Every card is compared on one model, Qwen3.8 27B at 2-bit, the size and bit class that most cards here have plain runs of. A card with no run of that model shows none rather than a figure from a different model.
| GPU | VRAM | Reports | Median speed, Qwen3.8 27B 2-bit |
|---|---|---|---|
| NVIDIA RTX 5090nvidia | 32GB | 4 | no Qwen3.8 27B 2-bit run reported |
| NVIDIA RTX 3090nvidia | 24GB | 2 | no Qwen3.8 27B 2-bit run reported |
| Intel Arc B580 12GBintel | 12GB | 1 | Gen 0.7 t/s |
| NVIDIA RTX 3060 12GBnvidia | 12GB | 1 | no Qwen3.8 27B 2-bit run reported |
| NVIDIA RTX 4060 Laptop 8GBnvidia | 8GB | 1 | no Qwen3.8 27B 2-bit run reported |
FreeToken against other engines
Pairs of plain runs on the same card, of the same model size at the same bit class: one device, one request, no speculative decoding, the whole model in memory. Context length and build still differ between the two sides, and each side shows its own.
No matched pair yet. No card has plain runs of one model size at one bit class on FreeToken and on another engine, so llamaperf can't say how it compares on speed. Add a run.
Top models on FreeToken
Frequently asked
Is FreeToken faster than other engines?
llamaperf has no matched comparison for FreeToken yet: no card has plain runs of the same model size at the same bit class on FreeToken and on another engine. Speed claims about engines need that pairing, so this page doesn't make one.