Model benchmarks

qwen3.6-14b-a3b-fablevibes local LLM performance

As of August 2026, qwen3.6-14b-a3b-fablevibes runs at up to 95.3 tok/s for local inference (best of 3 community benchmark runs across 2 GPUs).

LM Studio
ShareRedditX

Model size

14B

Peak speed

95.3 tok/s

Average speed

71.4 tok/s

Min memory

8.7 GB

Max context

16.384 tokens

Avg quality

50.5

Benchmark runs

3

GPUs tested

2

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 3 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)
Overall45.150.557.6
Agent Workflow57.962.266.6
Code Generation32.439.044.2
Role Play & Narrative24.440.261.1
Research & Analysis53.160.770.6

Performance by hardware and tool

Every hardware/tool/quantization combination qwen3.6-14b-a3b-fablevibes has been benchmarked on, ranked by peak token generation speed. Last updated August 2026.

HardwareToolQuantBest tok/sAvg tok/sMemoryContextQualityRuns
AMD Radeon RX 6800 XTLM Studio95.3 tok/s84.4 tok/s8.7 GB8.192 tokens46.52
NVIDIA GeForce RTX 4070 Ti SUPERLM Studio45.2 tok/s45.2 tok/s14.5 GB16.384 tokens58.61

Benchmark runs

All 3 qwen3.6-14b-a3b-fablevibes runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.

Frequently asked questions

Is qwen3.6-14b-a3b-fablevibes good for coding?
In our benchmarks, qwen3.6-14b-a3b-fablevibes scores 39.0/100 for coding. It runs at about 71.4 tok/s, so if you want more speed, unsloth/Qwen3.8-27B-GGUF:IQ3_S is faster (~104.3 tok/s) and still scores well for coding (83.0/100).
Is qwen3.6-14b-a3b-fablevibes good for agentic (tool-using) tasks?
In our benchmarks, qwen3.6-14b-a3b-fablevibes scores 62.2/100 for agentic workflows. It runs at about 71.4 tok/s, so if you want more speed, Tiel-Coder-35B-A3B-Q4_K_S is faster (~165.0 tok/s) and still scores well for agentic workflows (89.3/100).
How fast is qwen3.6-14b-a3b-fablevibes for local inference?
Across 3 community benchmark runs, qwen3.6-14b-a3b-fablevibes reaches up to 95.3 tok/s and averages 71.4 tok/s, with the fastest results on AMD Radeon RX 6800 XT.
How much memory does qwen3.6-14b-a3b-fablevibes need?
The leanest observed configuration used about 8.7 GB of memory.
Which tools have been used to run qwen3.6-14b-a3b-fablevibes?
Benchmarks were submitted using LM Studio. Results are community-contributed and updated as new runs arrive.