Benchmark result

qwen3_coder_next on NVIDIA GeForce RTX 3080 — 31.1 tok/s

Measured with LM Studio on August 25, 2026.

What the model built

The coding scenario asks for a playable game in a single HTML file. This is exactly what the model returned, unedited; the frame only adds a line that tells this page how tall it is.

How this run compares

4th fastest of 4 runs of this model on NVIDIA GeForce RTX 3080 · median 31.2 tok/s.

sort

131.2tok/s80.6LM Studio
231.2tok/s78.9LM Studio
331.2tok/s77.4LM Studio
431.1tok/s83.6LM Studiothis run

Compare the nearest 5 →

qwen3_coder_next on other hardware

51.9tok/s76.6Apple M4 Max
34.1tok/s74.2Apple M5 Max · Q4_K_M
33.4tok/s75.1Apple M5 Max · Q4_K_M
29.4tok/s68.2Apple M5 Max · Q4_K_M
28.9tok/s78.2Apple M5 Max · Q4_K_M

Reproduce this run

toolLM Studio

modelqwen3_coder_next

context8,192

clientv0.4.43+97

Set those in LM Studio, then:

llm-benchmark benchmark --model "qwen3_coder_next" --tool "LM Studio"

The client prompts for context length and thinking mode, and for the KV cache dtype on Unsloth Studio. Sampling is left at the model default — the values above are what the tool reported using, not overrides the client sent.