← Leaderboard
Qwen3.6-35B-A3B MLX 8bit Uniform (lmstudio-community)
Retired Rank #44 of 51 · 2/8 GAUNTLET progress
Hollow markers & dashed spokes: axis not yet scored
G
88
A
—
U
—
N
—
T
n/d
L
—
E
—
T
25.3
Specification
- Parameters
- 35B total / 3B active per token
- Architecture
- qwen3_5_moe
- Size on disk
- 37.75 GB
- Quantization
- 8bit uniform
- Format
- MLX
- Runtime
- LM Studio
- Class
- Standard
- Reasoning (CoT)
- Yes — emits reasoning tokens
- Internal ID
- M9
- Mean speed
- 88.8 tok/s across suites
- Stall census
- 1 stall in 13 observed tests (7.7%)
- Reasoning appetite
- 3,254 tokens mean · 7,730 max
- Model card
- lmstudio.ai
Suite results
Suite Score Avg / 20 Tests n tok/s Run
General capability (13-task real-workload suite) 1a 228.8 / 260 17.6 13/13 n=2 88.8 run 2026-09-02
| Test | Category | Score | tok/s | |
|---|---|---|---|---|
| A1 | Dev/Ops Scripting | 14/20 | 89.5 | |
| A2 | Dev/Ops Scripting | 17/20 | 89.4 | |
| A3 | Dev/Ops Scripting | 20/20 | 89.3 | |
| A4 | Dev/Ops Scripting | 19/20 | 91.2 | |
| A5 | Dev/Ops Scripting | 20/20 | 91.3 | |
| A6 | Dev/Ops Scripting | 20/20 | 91.1 | |
| B1 | Document Processing | 20/20 | 88.9 | |
| B2 | Document Processing | 20/20 | 87.3 | |
| B3 | Document Processing | 0/20 | 82.9 | |
| C1 | Content Production | 20/20 | 83.9 | |
| C2 | Content Production | 19/20 | 82.6 | |
| C3 | Content Production | 20/20 | 84.2 | |
| C4 | Content Production | 20/20 | 85.9 |
Category mean Dev/Ops Scripting 18.3Document Processing 13.3Content Production 19.8
Each test is LLM-judged 0–20 against a fixed rubric; suite max = tests × 20. An ADJ chip marks a cell carrying a subjective adjustment from the maintainer; the judged score is the one shown, is unchanged, and remains what the axes, ranks and tie bands are computed from. Hover for the reason. Rows with a ▶ expand to the per-test breakdown. Where fewer tests ran than the suite total, unrun tests count as zero toward the axis score. See methodology.