3 builds of Qwen 3.5 0.8B measured on real machines — different packagers, precisions and engines, each a different file with its own numbers. The best each engine proved is first; every build has its own page with the machines and signatures behind it.
proved by at least one build: Tool calling · Structured output · Extended thinking · Writes code · Follows instructions
| build | publisher | precision | engine | whose work | context proved | best tok/s | machines |
|---|---|---|---|---|---|---|---|
| hf.co/unsloth/Qwen3.5-0.8B-GGUF:Q4_K_M | unsloth | Q4_K_M | ollama | quantizer | — | 17.0 | 2 |
| hf.co/unsloth/Qwen3.5-0.8B-GGUF:Q4_0 | unsloth | Q4_0 | ollama | quantizer | — | 16.6 | 1 |
| hf.co/unsloth/Qwen3.5-0.8B-GGUF:Q8_0 | unsloth | Q8_0 | ollama | quantizer | — | 12.7 | 1 |