← all models

MODEL RESULTS — QWEN3.8-27B

qwen3.8-flash-next-iq3_xxs · server http://strata · run 2026-10-12 19:16:07 → 2026-10-12 19:16:08

OVERALL

Composite: 9.1 ✓ 8/8 tests
Speed (median): 34.1 tok/s
Thinking cost: 0 tokens · ≈0s · starved: none

ENVIRONMENT & LOAD (verified)

THE BENCHMARK REFERENCE ROW — Qwen3.8-27B Q4_K_M, composite 9.1 (8/8), 34.1 tok/s. Every other model on this page is ranked against it. Run 2026-10-04 on box strata before its retirement; some per-test token data was lost in the layout migration, and the exact context/GPU-residency of this run was never recorded.
Model fileQwen3.8-27B-Q4_K_M.gguf (manifest path: C:\Users\Admin\.lmstudio\models\lmstudio-community\Qwen3.8-27B-GGUF\)
Server evidenceswift-1.5-iq2-s env.json (same box, same week): 'llama-server PID 2040 on the Arc B70 is still running the benchmark's OWN reference model (Qwen3.8-27B Q4_K_M) on '
Manifest server fieldlost — reads 'migrated-from-old-layout'
Run window2026-10-04 18:20 → 19:04
Token datapartial — breakout elapsed 783 s recorded, but total/reasoning tokens are None (old-layout migration)
Scorescomposite 9.1, 8/8 tests — the top score on the board as of 2026-10-12
Observations
· [assistant] 34.1 tok/s on 27B Q4_K_M is far above what a 12 GB 3080 Ti could do with 16 GB of weights — consistent with the Arc B70 record, not the 3080 Ti. (inference from swift env record + speed)
Still to verify
· Exact context window and GPU residency of this run were never recorded.
· Box strata retired Oct 2026 — this row can no longer be re-verified or re-run; record frozen as of migration.
· 2026-10-13 GPU-attribution caveat: the same physical box (now strata) serves on its NVIDIA RTX 3080 Ti per GET /metrics (hardware_static.gpu_name), and its Intel Arc (vendor 0x8086, driver xe) is installed but unused for serving today. The 'Arc B70' attribution for this 2026-10-04 run rests only on the swift-1.5-iq2-s env record and is therefore less certain than it used to read.

PER-TEST BREAKDOWN

TestScoreCheck / detailContent tokThink tok≈Think sElapsed stok/sOutput
speed—18906.8627.54speed.txt ↗
fib10 ·reviewfib(0..10) correct, fib(50)=12586269025270019.0214.19fib.txt ↗
reasoning9 ·review9303.1829.24reasoning.txt ↗
constraints10 ·review3 sentence(s) found (target 3)17305.2832.79constraints.txt ↗
json10 ·reviewvalid JSON, all keys + types correct10803.0535.46json.txt ↗
creative9 ·review600017.6134.07creative.txt ↗
breakout9 ·review320000783.340.85breakout.html ↗
delta8 ·review320000873.9636.62delta.html ↗
fractal9 ·review320000933.5634.28fractal.html ↗

GENERATED GAMES

breakout
breakout · canvas painted · score 9 · 0 render errs
Open breakout.html ↗screenshot
delta
delta · canvas painted · score 8 · 0 render errs
Open delta.html ↗screenshot
fractal
fractal · canvas painted · score 9 · 0 render errs
Open fractal.html ↗screenshot

RAW TEXT / CODE OUTPUTS

speed.txt (140 bytes, full file beside this page)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
fib.txt (152 bytes, full file beside this page)
def fib(n):
    if n < 0:
        raise ValueError("n must be non-negative")
    a, b = 0, 1
    for _ in range(n):
        a, b = b, a + b
    return a
reasoning.txt (89 bytes, full file beside this page)
“All but 9 die” means 9 sheep do not die, so they are the ones left.

**Final number: 9**
constraints.txt (250 bytes, full file beside this page)
Rain taps a steady rhythm on the windowpane, blurring the city into soft gray shapes. The air smells of wet stone and distant thunder, while puddles collect like scattered mirrors. Under a tilted umbrella, the day feels quiet, cool, and gently alive.
json.txt (66 bytes, full file beside this page)
{"name":"Alice","age":30,"hobbies":["reading","hiking","cooking"]}
creative.txt (0 bytes, full file beside this page)
Genuine model outputs — files saved verbatim under qwen3.8-27b/ (source of truth in tests/<name>/result.json). Game links open the raw HTML the model produced.