Model benchmarks

ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS local LLM performance

As of November 2026, ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS runs at up to 54.2 tok/s for local inference (best of 8 community benchmark runs across 1 GPU).

OpenAI-compatible
ShareRedditX

Model size

177B

Peak speed

54.2 tok/s

Average speed

48.1 tok/s

Avg PP

228.7 tok/s

Min memory

31.9 GB

Max context

262,144 tokens

Avg output / run

41,056 tokens

Avg runtime / run

15m 30s

Avg quality

84.9

Benchmark runs

8

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 8 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.

TaskP5 (low)AvgP95 (high)
Overall80.984.988.7
Agent Workflow82.287.892.2
Code Generation66.674.482.2
Role Play & Narrative80.590.195.5
Research & Analysis82.787.291.1

Performance by hardware and tool

Every hardware/tool/quantization combination ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS has been benchmarked on, ranked by peak token generation speed. Last updated November 2026.

HardwareToolQuantBest tok/sAvg tok/sMemoryContextQualityRuns
Apple M4 ProOpenAI-compatible—54.2 tok/s48.1 tok/s31.9 GB262,144 tokens84.98

Benchmark runs

All 8 ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.

Frequently asked questions

Is ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS good for coding?
In our benchmarks, ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS scores 74.4/100 for coding. It runs at about 48.1 tok/s, so if you want more speed, turbo is faster (~124.1 tok/s) and still scores well for coding (85.3/100).
Is ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS good for agentic (tool-using) tasks?
In our benchmarks, ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS scores 87.8/100 for agentic workflows. It runs at about 48.1 tok/s, so if you want more speed, granite-4.2-8b-Q4_K_L is faster (~71.6 tok/s) and still scores well for agentic workflows (92.9/100).
How fast is ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS for local inference?
Across 8 community benchmark runs, ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS reaches up to 54.2 tok/s and averages 48.1 tok/s, with the fastest results on Apple M4 Pro.
How much memory does ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS need?
The leanest observed configuration used about 31.9 GB of memory.
Which tools have been used to run ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF:IQ3_XXS?
Benchmarks were submitted using OpenAI-compatible. Results are community-contributed and updated as new runs arrive.