AI judge results

Each judge sees two anonymous answers to the same native English writing prompt. It picks the stronger text or declares a tie, without seeing the model names.

A judge may also evaluate a pair containing a response from its own model; the answer identities remain hidden while it decides.

The tables add up every validated, published duel for each judge: wins, losses, ties and total comparisons. Blinding keeps each verdict focused on the writing rather than the brand behind it.

Ranks are ordered first by total wins, not by win rate or opponent strength; when coverage differs, read wins, losses and ties together with comparisons.

For the selected writing categories, ranks are assigned before publisher, size, model-status, generation-mode and access filters hide models, so gaps can remain in the visible numbering.

AI judges make it possible to compare many model pairs with a consistent protocol. They do not replace readers: human votes are counted separately in the human ranking, revealing where people and machines agree or disagree.

Latest included verdict: September 19, 2026 at 10:47 p.m. CEST

Claude Fable 5

1,084 validated duel verdicts

Results from Claude Fable 5
Rank	Model	Wins	Losses	Ties	Comparisons
2	MiniMax M3 💡 enabled	44	39	8	91
3	Claude Opus 5 💡 adaptive high	24	10	11	45
4	Qwen3.7 Plus 💡 enabled	15	53	4	72
7	MiMo V2.5 Pro 💡 enabled	14	53	5	72
8	DeepSeek V4 Pro 💡 enabled	13	51	7	71
9	Kimi K3 💡 enabled	9	16	20	45
13	MiMo V2.5 💡 enabled	7	62	3	72
14	Gemma 4 31B QAT-Q4_0 💡 on* 🏠	6	29	1	36
15	Gemma 4 12B QAT-Q4_0 💡 on* 🏠	4	23	1	28
17	Gemma 4 12B Q8_0 ↯ 🏠	3	10	4	17
18	Qwen3.6 35B A3B Q8_0 💡 on* 🏠	3	31	3	37
19	GPT-5.6 Luna 💡 high	3	39	3	45
20	GPT-5.6 Terra 💡 high	3	39	3	45
22	Gemma 4 12B Q8_0 💡 on* 🏠	2	30	6	38
23	GPT-5.6 Sol 💡 high	2	36	6	44
24	Gemma 4 31B QAT-Q4_0 ↯ 🏠	1	4	1	6
25	Qwen3.6 35B A3B Q8_0 ↯ 🏠	1	11	5	17
26	Gemma 4 31B Q8_0 💡 on* 🏠	1	28	4	33
27	Qwen3.6 35B A3B Q4_K_M 💡 on* 🏠	1	30	3	34
28	Mistral Medium 3.5 128B Q4_K_M ↯ 🏠	0	8	0	8
29	Gemma 4 26B A4B Q8_0 💡 on* 🏠	0	22	2	24
30	Gemma 4 26B A4B QAT-Q4_0 💡 on* 🏠	0	31	0	31

Claude Fable 5.1

1,741 validated duel verdicts

Results from Claude Fable 5.1
Rank	Model	Wins	Losses	Ties	Comparisons
1	Claude Fable 5.1 💡 adaptive high	147	17	25	189
2	Claude Opus 5 💡 adaptive high	118	7	12	137
3	GLM 5.3 💡 max	102	34	31	167
4	Kimi K3 💡 enabled	96	22	19	137
5	Gemini 3.8 Flash 💡 high	79	64	35	178
6	Muse Spark 1.3 💡 high	76	79	33	188
7	Claude Sonnet 5 💡 adaptive high	69	51	22	142
8	Hy3 💡 high	59	72	32	163
9	GLM 5.3 Flash 💡 high	55	45	21	121
10	MiniMax M3 💡 enabled	51	66	18	135
11	GPT-6 Astra 💡 high	44	60	24	128
12	Grok 4.6 💡 high	40	10	5	55
13	GPT-5.6 Sol 💡 high	38	72	25	135
14	Seed 2.1 Turbo ↯	35	15	1	51
15	Aion 3.0 💡 mandatory	34	6	6	46
16	DeepSeek V4.1 Flash 💡 high	31	70	9	110
17	Inkling 💡 high	27	10	13	50
18	Qwen3.8 Max 💡 xhigh	25	17	13	55
19	Qwen3.8 27B Q4_K_M 💡 xhigh** 🏠	20	3	8	31
20	Qwen3.8 27B Q6_K ↯ 🏠	18	8	7	33
21	Gemma 4 26B A4B Q6_K ↯ 🏠	16	5	7	28
22	Granite 4.2 30B Q6_K 💡 full 🏠	15	35	8	58
23	Gemma 4 31B Q6_K 💡 on* 🏠	14	9	4	27
24	Qwen3.8 27B Q4_K_M ↯ 🏠	13	7	5	25
26	Mistral Medium 3.5 💡 high	13	25	13	51
27	Gemma 4 E4B Q6_K ↯ 🏠	12	13	2	27
28	Qwen3.8 27B Q6_K 💡 xhigh** 🏠	11	7	5	23
29	Gemma 4 26B A4B QAT-Q4_0 ↯ 🏠	11	14	6	31
30	Gemma 4 31B Q6_K ↯ 🏠	11	16	2	29
32	Gemma 4 12B Q6_K 💡 on* 🏠	10	5	7	22
33	Qwen3.8 27B Q8_0 ↯ 🏠	10	9	5	24
34	Gemma 4 26B A4B Q6_K 💡 on* 🏠	9	11	8	28
35	Gemma 4 E4B Q4_K_M ↯ 🏠	9	16	4	29
36	Gemma 4 E4B Q6_K 💡 on* 🏠	9	17	3	29
37	Gemma 4 26B A4B Q8_0 ↯ 🏠	8	6	13	27
38	DeepSeek V4 Pro 💡 enabled	8	9	2	19
39	Qwen3.8 27B Q8_0 💡 xhigh** 🏠	8	22	2	32
41	Gemma 4 E4B Q4_K_M 💡 on* 🏠	7	18	2	27
42	Qwen3.6 35B A3B Q4_K_M ↯ 🏠	7	21	2	30
43	Muse Glimmer 30B K-Quant Dynamic 💡 high 🏠	6	12	4	22
44	MiMo V2.5 💡 enabled	6	13	1	20
45	MiMo V2.5 Pro 💡 enabled	5	6	9	20
46	Gemma 4 E4B Q8_0 ↯ 🏠	5	7	3	15
47	Gemma 4 31B Q8_0 ↯ 🏠	5	8	14	27
48	Mistral Medium 3.5 128B Q4_K_M ↯ 🏠	5	11	3	19
49	Gemma 4 12B Q8_0 ↯ 🏠	5	12	3	20
50	Muse Glimmer 30B K-Quant 17GB 💡 high 🏠	5	13	4	22
51	Nemotron 3 Super 120B A12B Q4_K_M 💡 full 🏠	5	20	1	26
52	Gemma 4 12B QAT-Q4_0 ↯ 🏠	5	26	0	31
53	Nemotron 3 Nano Omni 30B A3B Q8_0 💡 on* 🏠	4	23	0	27
54	Gemma 4 E2B Q4_K_M ↯ 🏠	4	31	0	35
55	GPT-5.6 Terra 💡 high	3	13	5	21
56	Qwen3.6 35B A3B Q8_0 ↯ 🏠	2	19	0	21
57	Gemma 4 12B Q6_K ↯ 🏠	2	22	0	24
58	Hy4 Preview 💡 high	1	2	1	4
59	Qwen3.8 Flash 💡 high	1	2	1	4
60	Granite 4.2 8B Q8_0 💡 full 🏠	1	6	0	7
61	Gemma 4 31B QAT-Q4_0 ↯ 🏠	1	14	6	21
62	Qwen3.7 Plus 💡 enabled	1	18	1	20
63	GPT-5.6 Luna 💡 high	1	19	1	21
64	Nemotron 3 Nano Omni 30B A3B Q4_K_M 💡 on* 🏠	1	24	0	25
65	Gemma 4 E2B Q6_K ↯ 🏠	1	30	1	32
66	Gemma 4 31B Q8_0 💡 on* 🏠	0	1	2	3
67	Gemma 4 12B QAT-Q4_0 💡 on* 🏠	0	2	1	3
68	Gemma 4 26B A4B Q8_0 💡 on* 🏠	0	2	1	3
69	Gemma 4 26B A4B QAT-Q4_0 💡 on* 🏠	0	2	1	3
70	Gemma 4 31B QAT-Q4_0 💡 on* 🏠	0	2	1	3
71	Qwen3.6 35B A3B Q4_K_M 💡 on* 🏠	0	2	1	3
72	Qwen3.6 35B A3B Q8_0 💡 on* 🏠	0	2	1	3
73	Gemma 4 E2B Q8_0 💡 on* 🏠	0	3	1	4
74	Gemma 4 12B Q8_0 💡 on* 🏠	0	3	0	3
75	Granite 4.2 30B Q8_0 💡 full 🏠	0	4	3	7
76	Gemma 4 E2B Q4_K_M 💡 on* 🏠	0	4	0	4
77	Granite 4.2 30B Q4_K_M 💡 full 🏠	0	7	0	7
78	Granite 4.2 8B Q4_K_M 💡 full 🏠	0	7	0	7
79	Gemma 4 E4B Q8_0 💡 on* 🏠	0	13	0	13
80	Gemma 4 E2B Q6_K 💡 on* 🏠	0	23	0	23
81	Gemma 4 E2B Q8_0 ↯ 🏠	0	28	2	30

Claude Opus 4.8

843 validated duel verdicts

Results from Claude Opus 4.8
Rank	Model	Wins	Losses	Ties	Comparisons
2	Qwen3.6 35B A3B Q4_K_M 💡 on* 🏠	78	68	5	151
3	MiniMax M3 💡 enabled	75	7	0	82
4	Gemma 4 31B QAT-Q4_0 💡 on* 🏠	62	76	9	147
5	Gemma 4 26B A4B QAT-Q4_0 💡 on* 🏠	56	106	9	171
6	Gemma 4 31B Q8_0 💡 on* 🏠	38	93	7	138
9	MiMo V2.5 Pro 💡 enabled	36	20	0	56
11	DeepSeek V4 Pro 💡 enabled	35	18	3	56
14	Qwen3.6 35B A3B Q8_0 💡 on* 🏠	29	34	2	65
15	MiMo V2.5 💡 enabled	28	27	0	55
16	Qwen3.7 Plus 💡 enabled	27	26	4	57
17	Gemma 4 26B A4B Q8_0 💡 on* 🏠	27	48	3	78
18	Gemma 4 12B Q8_0 💡 on* 🏠	21	42	1	64
19	Gemma 4 12B QAT-Q4_0 💡 on* 🏠	19	66	3	88
21	GPT-5.6 Sol 💡 high	5	1	1	7
22	Mistral Medium 3.5 128B Q4_K_M ↯ 🏠	5	6	0	11
23	Gemma 4 12B Q8_0 ↯ 🏠	5	9	0	14
24	Qwen3.6 35B A3B Q8_0 ↯ 🏠	5	10	0	15
25	Gemma 4 31B QAT-Q4_0 ↯ 🏠	0	3	0	3

Claude Opus 5

28,248 validated duel verdicts

Results from Claude Opus 5
Rank	Model	Wins	Losses	Ties	Comparisons
1	Claude Opus 5 💡 adaptive high	741	40	15	796
3	Kimi K3 💡 enabled	728	48	20	796
4	Claude Sonnet 5 💡 adaptive high	659	87	48	794
5	MiniMax M3 💡 enabled	632	133	29	794
6	Muse Glimmer 30B K-Quant Dynamic 💡 high 🏠	587	143	57	787
7	GPT-6 Astra 💡 high	579	52	66	697
8	Claude Fable 5.1 💡 adaptive high	575	30	14	619
10	GPT-5.6 Sol 💡 high	574	178	47	799
11	GLM 5.3 💡 max	538	46	55	639
12	DeepSeek V4 Pro 💡 enabled	526	240	32	798
13	GLM 5.3 Flash 💡 high	521	80	38	639
14	DeepSeek V4.1 Flash 💡 high	505	120	55	680
15	Muse Glimmer 30B K-Quant 17GB 💡 high 🏠	502	235	45	782
16	GPT-5.6 Terra 💡 high	486	256	56	798
17	Hy3 💡 high	468	137	50	655
18	Aion 3.0 💡 mandatory	459	155	43	657
19	Gemini 3.8 Flash 💡 high	453	123	51	627
20	Muse Spark 1.3 💡 high	450	131	33	614
21	Qwen3.8 Max 💡 xhigh	437	138	72	647
22	GPT-5.6 Luna 💡 high	435	315	46	796
23	Qwen3.6 35B A3B Q8_0 💡 on* 🏠	418	390	54	862
24	Hy4 Preview 💡 high	403	183	54	640
26	Grok 4.6 💡 high	388	234	21	643
27	Qwen3.8 Flash 💡 high	384	232	25	641
29	Inkling 💡 high	371	285	49	705
30	Gemma 4 31B Q8_0 💡 on* 🏠	370	367	67	804
31	Qwen3.8 27B Q4_K_M 💡 xhigh** 🏠	361	168	43	572
32	Gemma 4 31B QAT-Q4_0 💡 on* 🏠	355	312	51	718
33	MiMo V2.5 Pro 💡 enabled	336	307	42	685
34	Qwen3.8 27B Q8_0 💡 xhigh** 🏠	335	176	37	548
35	MiMo V2.5 💡 enabled	329	325	29	683
36	Qwen3.8 27B Q6_K 💡 xhigh** 🏠	323	175	55	553
37	Gemma 4 26B A4B Q6_K ↯ 🏠	320	326	61	707
38	Gemma 4 31B QAT-Q4_0 ↯ 🏠	320	405	61	786
39	Gemma 4 31B Q8_0 ↯ 🏠	306	328	62	696
40	Gemma 4 31B Q6_K ↯ 🏠	292	363	62	717
41	Gemma 4 26B A4B Q6_K 💡 on* 🏠	291	222	64	577
42	Gemma 4 31B Q6_K 💡 on* 🏠	291	222	52	565
43	Qwen3.7 Plus 💡 enabled	290	364	38	692
47	Qwen3.6 35B A3B Q4_K_M 💡 on* 🏠	264	383	35	682
48	Gemma 4 26B A4B QAT-Q4_0 💡 on* 🏠	259	386	57	702
49	Gemma 4 26B A4B Q8_0 💡 on* 🏠	254	409	43	706
50	Gemma 4 12B Q8_0 💡 on* 🏠	249	411	50	710
51	Gemma 4 12B Q8_0 ↯ 🏠	242	413	45	700
52	Gemma 4 12B QAT-Q4_0 💡 on* 🏠	236	418	63	717
54	Seed 2.1 Turbo ↯	229	245	17	491
55	Qwen3.8 27B Q6_K ↯ 🏠	225	476	16	717
56	Nemotron 3 Super 120B A12B Q4_K_M 💡 full 🏠	214	415	33	662
58	Qwen3.6 35B A3B Q8_0 ↯ 🏠	197	527	20	744
59	Mistral Medium 3.5 128B Q4_K_M ↯ 🏠	196	459	29	684
60	Gemma 4 12B Q6_K 💡 on* 🏠	184	269	41	494
61	Gemma 4 26B A4B QAT-Q4_0 ↯ 🏠	182	302	26	510
62	Ling 3.0 Flash Q4_K_M ↯ 🏠	170	176	21	367
63	Qwen3.8 27B Q8_0 ↯ 🏠	162	545	12	719
66	Qwen3.8 27B Q4_K_M ↯ 🏠	145	549	23	717
67	Gemma 4 12B QAT-Q4_0 ↯ 🏠	144	334	23	501
68	Gemma 4 26B A4B Q8_0 ↯ 🏠	142	336	29	507
70	Qwen3.6 35B A3B Q4_K_M ↯ 🏠	129	365	10	504
71	Gemma 4 E2B Q8_0 💡 on* 🏠	129	533	36	698
72	Gemma 4 12B Q6_K ↯ 🏠	120	352	27	499
73	Nemotron 3 Nano Omni 30B A3B Q4_K_M 💡 on* 🏠	116	558	10	684
74	Mistral Medium 3.5 💡 high	113	335	18	466
75	Ling 3.0 Flash Q4_K_M 💡 high 🏠	109	241	10	360
76	Gemma 4 E4B Q8_0 💡 on* 🏠	103	580	9	692
77	Gemma 4 E4B Q8_0 ↯ 🏠	97	585	14	696
78	Granite 4.2 30B Q6_K 💡 full 🏠	91	287	19	397
79	Granite 4.2 30B Q8_0 💡 full 🏠	88	230	24	342
80	Nemotron 3 Nano Omni 30B A3B Q8_0 💡 on* 🏠	82	362	15	459
81	Gemma 4 E4B Q4_K_M ↯ 🏠	81	420	4	505
82	Granite 4.2 8B Q8_0 💡 full 🏠	79	371	14	464
83	Gemma 4 E4B Q6_K ↯ 🏠	77	415	7	499
84	Gemma 4 E4B Q6_K 💡 on* 🏠	68	417	5	490
85	Gemma 4 E2B Q4_K_M 💡 on* 🏠	64	612	22	698
86	Gemma 4 E2B Q8_0 ↯ 🏠	63	434	4	501
87	Granite 4.2 8B Q4_K_M 💡 full 🏠	59	358	21	438
88	Gemma 4 E2B Q6_K 💡 on* 🏠	55	414	24	493
89	Gemma 4 E4B Q4_K_M 💡 on* 🏠	53	421	8	482
90	Granite 4.2 30B Q4_K_M 💡 full 🏠	49	268	19	336
91	Gemma 4 E2B Q6_K ↯ 🏠	33	468	2	503
92	Gemma 4 E2B Q4_K_M ↯ 🏠	32	467	4	503

GPT-5.6 Sol

24,797 validated duel verdicts

Results from GPT-5.6 Sol
Rank	Model	Wins	Losses	Ties	Comparisons
2	GPT-5.6 Sol 💡 high	866	114	12	992
3	Kimi K3 💡 enabled	543	78	4	625
5	Claude Opus 5 💡 adaptive high	505	120	0	625
6	GPT-5.6 Terra 💡 high	473	154	5	632
7	Claude Sonnet 5 💡 adaptive high	468	160	2	630
9	MiniMax M3 💡 enabled	452	218	0	670
10	GPT-5.6 Luna 💡 high	448	177	7	632
11	GPT-6 Astra 💡 high	423	18	2	443
12	DeepSeek V4 Pro 💡 enabled	415	407	5	827
13	Gemma 4 31B Q8_0 💡 on* 🏠	405	300	8	713
14	Muse Glimmer 30B K-Quant Dynamic 💡 high 🏠	396	197	6	599
15	Qwen3.7 Plus 💡 enabled	389	284	9	682
16	GLM 5.3 💡 max	381	60	8	449
17	GLM 5.3 Flash 💡 high	372	173	1	546
18	Hy4 Preview 💡 high	359	172	6	537
19	Claude Fable 5.1 💡 adaptive high	354	82	2	438
20	Qwen3.8 Max 💡 xhigh	348	92	21	461
21	Muse Spark 1.3 💡 high	346	83	2	431
22	Gemma 4 31B QAT-Q4_0 💡 on* 🏠	343	370	4	717
23	Qwen3.6 35B A3B Q8_0 💡 on* 🏠	339	329	4	672
24	DeepSeek V4.1 Flash 💡 high	338	88	3	429
25	Gemma 4 31B QAT-Q4_0 ↯ 🏠	336	309	9	654
26	Hy3 💡 high	335	111	2	448
27	Aion 3.0 💡 mandatory	322	144	1	467
28	Qwen3.8 Flash 💡 high	321	213	7	541
29	Muse Glimmer 30B K-Quant 17GB 💡 high 🏠	320	281	2	603
30	Gemma 4 12B QAT-Q4_0 💡 on* 🏠	318	348	6	672
31	Grok 4.6 💡 high	311	149	2	462
32	Gemini 3.8 Flash 💡 high	310	123	6	439
33	MiMo V2.5 Pro 💡 enabled	310	364	7	681
34	Gemma 4 26B A4B Q8_0 💡 on* 🏠	310	381	3	694
36	Qwen3.6 35B A3B Q4_K_M 💡 on* 🏠	305	405	3	713
37	Gemma 4 26B A4B QAT-Q4_0 💡 on* 🏠	301	424	4	729
38	MiMo V2.5 💡 enabled	289	393	0	682
39	Gemma 4 12B Q8_0 💡 on* 🏠	284	381	0	665
42	Qwen3.8 27B Q8_0 💡 xhigh** 🏠	272	180	2	454
43	Gemma 4 12B Q8_0 ↯ 🏠	264	433	4	701
45	Qwen3.8 27B Q6_K 💡 xhigh** 🏠	254	194	0	448
46	Gemma 4 26B A4B Q6_K 💡 on* 🏠	250	178	3	431
49	Gemma 4 31B Q6_K 💡 on* 🏠	244	177	5	426
50	Qwen3.8 27B Q4_K_M 💡 xhigh** 🏠	241	178	5	424
51	Gemma 4 26B A4B Q6_K ↯ 🏠	238	224	5	467
52	Ling 3.0 Flash Q4_K_M ↯ 🏠	236	240	1	477
53	Gemma 4 31B Q6_K ↯ 🏠	235	212	1	448
54	Granite 4.2 30B Q8_0 💡 full 🏠	228	277	2	507
56	Gemma 4 31B Q8_0 ↯ 🏠	217	243	4	464
57	Nemotron 3 Super 120B A12B Q4_K_M 💡 full 🏠	216	386	4	606
58	Gemma 4 12B Q6_K 💡 on* 🏠	203	212	0	415
60	Inkling 💡 high	180	269	0	449
61	Gemma 4 E2B Q8_0 💡 on* 🏠	177	424	1	602
62	Gemma 4 26B A4B QAT-Q4_0 ↯ 🏠	176	237	3	416
63	Ling 3.0 Flash Q4_K_M 💡 high 🏠	175	302	0	477
64	Mistral Medium 3.5 128B Q4_K_M ↯ 🏠	161	704	5	870
65	Qwen3.6 35B A3B Q8_0 ↯ 🏠	159	546	0	705
66	Granite 4.2 8B Q8_0 💡 full 🏠	150	356	3	509
67	Gemma 4 12B QAT-Q4_0 ↯ 🏠	149	260	3	412
68	Nemotron 3 Nano Omni 30B A3B Q8_0 💡 on* 🏠	148	251	2	401
69	Mistral Medium 3.5 💡 high	145	260	2	407
70	Gemma 4 E2B Q4_K_M 💡 on* 🏠	142	459	0	601
71	Seed 2.1 Turbo ↯	137	278	0	415
72	Granite 4.2 30B Q6_K 💡 full 🏠	134	235	0	369
73	Granite 4.2 8B Q4_K_M 💡 full 🏠	134	388	1	523
74	Gemma 4 12B Q6_K ↯ 🏠	128	282	1	411
75	Gemma 4 26B A4B Q8_0 ↯ 🏠	123	281	3	407
76	Gemma 4 E4B Q8_0 💡 on* 🏠	114	475	1	590
77	Gemma 4 E4B Q8_0 ↯ 🏠	113	473	0	586
78	Nemotron 3 Nano Omni 30B A3B Q4_K_M 💡 on* 🏠	109	502	0	611
79	Qwen3.8 27B Q8_0 ↯ 🏠	108	336	0	444
81	Gemma 4 E4B Q6_K 💡 on* 🏠	107	290	0	397
82	Granite 4.2 30B Q4_K_M 💡 full 🏠	103	399	1	503
83	Qwen3.8 27B Q6_K ↯ 🏠	101	345	0	446
84	Gemma 4 E2B Q6_K 💡 on* 🏠	100	315	0	415
85	Qwen3.6 35B A3B Q4_K_M ↯ 🏠	97	306	0	403
86	Qwen3.8 27B Q4_K_M ↯ 🏠	92	350	0	442
87	Gemma 4 E4B Q4_K_M ↯ 🏠	91	311	0	402
88	Gemma 4 E4B Q6_K ↯ 🏠	82	322	0	404
89	Gemma 4 E4B Q4_K_M 💡 on* 🏠	75	326	0	401
90	Gemma 4 E2B Q8_0 ↯ 🏠	74	325	0	399
91	Gemma 4 E2B Q4_K_M ↯ 🏠	69	330	0	399
92	Gemma 4 E2B Q6_K ↯ 🏠	46	352	0	398

Muse Spark 1.3

6,325 validated duel verdicts

Results from Muse Spark 1.3
Rank	Model	Wins	Losses	Ties	Comparisons
1	GLM 5.3 Flash 💡 high	169	16	12	197
2	Muse Spark 1.3 💡 high	134	22	16	172
3	Hy3 💡 high	132	13	17	162
4	Claude Fable 5.1 💡 adaptive high	131	26	4	161
5	Gemini 3.8 Flash 💡 high	125	25	22	172
6	Hy4 Preview 💡 high	122	50	24	196
7	Granite 4.2 8B Q4_K_M 💡 full 🏠	119	91	11	221
8	GLM 5.3 💡 max	112	25	6	143
10	GPT-6 Astra 💡 high	110	4	26	140
11	Claude Opus 5 💡 adaptive high	109	12	3	124
12	Inkling 💡 high	109	43	10	162
13	Kimi K3 💡 enabled	108	6	10	124
14	Claude Sonnet 5 💡 adaptive high	105	13	6	124
15	MiniMax M3 💡 enabled	104	18	6	128
16	GPT-5.6 Sol 💡 high	103	10	18	131
17	Qwen3.8 Max 💡 xhigh	103	46	11	160
18	Qwen3.8 Flash 💡 high	102	80	16	198
19	Grok 4.6 💡 high	101	44	15	160
20	Qwen3.8 27B Q6_K ↯ 🏠	97	60	7	164
21	DeepSeek V4 Pro 💡 enabled	93	48	8	149
22	Granite 4.2 8B Q8_0 💡 full 🏠	93	84	19	196
24	DeepSeek V4.1 Flash 💡 high	85	51	22	158
25	Gemma 4 31B QAT-Q4_0 💡 on* 🏠	84	46	26	156
26	Muse Glimmer 30B K-Quant 17GB 💡 high 🏠	83	29	17	129
27	Gemma 4 26B A4B Q6_K 💡 on* 🏠	81	45	17	143
28	Muse Glimmer 30B K-Quant Dynamic 💡 high 🏠	80	26	22	128
29	Gemma 4 31B Q6_K 💡 on* 🏠	78	55	19	152
30	Aion 3.0 💡 mandatory	78	59	23	160
31	Qwen3.8 27B Q4_K_M 💡 xhigh** 🏠	77	38	16	131
32	Gemma 4 26B A4B Q6_K ↯ 🏠	77	88	27	192
33	Qwen3.8 27B Q8_0 ↯ 🏠	75	93	5	173
34	Gemma 4 31B Q8_0 💡 on* 🏠	74	52	29	155
35	Gemma 4 12B Q6_K 💡 on* 🏠	73	58	34	165
36	Qwen3.8 27B Q8_0 💡 xhigh** 🏠	72	53	16	141
37	Granite 4.2 30B Q8_0 💡 full 🏠	72	139	13	224
38	Gemma 4 31B Q6_K ↯ 🏠	71	78	15	164
39	Gemma 4 12B QAT-Q4_0 💡 on* 🏠	69	46	29	144
40	Qwen3.6 35B A3B Q8_0 💡 on* 🏠	67	45	25	137
41	Gemma 4 12B Q8_0 💡 on* 🏠	67	59	25	151
42	Gemma 4 26B A4B Q8_0 💡 on* 🏠	66	59	25	150
44	GPT-5.6 Luna 💡 high	65	55	29	149
45	Qwen3.8 27B Q6_K 💡 xhigh** 🏠	64	53	22	139
46	Gemma 4 26B A4B QAT-Q4_0 ↯ 🏠	63	80	16	159
47	MiMo V2.5 💡 enabled	62	57	21	140
48	GPT-5.6 Terra 💡 high	62	70	15	147
49	Gemma 4 31B Q8_0 ↯ 🏠	62	77	21	160
50	Qwen3.8 27B Q4_K_M ↯ 🏠	61	86	26	173
51	Gemma 4 31B QAT-Q4_0 ↯ 🏠	59	52	30	141
52	Gemma 4 26B A4B QAT-Q4_0 💡 on* 🏠	59	71	21	151
53	MiMo V2.5 Pro 💡 enabled	59	73	8	140
54	Qwen3.6 35B A3B Q4_K_M ↯ 🏠	59	100	8	167
55	Qwen3.7 Plus 💡 enabled	57	70	16	143
56	Gemma 4 26B A4B Q8_0 ↯ 🏠	55	92	14	161
57	Seed 2.1 Turbo ↯	54	90	14	158
58	Qwen3.6 35B A3B Q8_0 ↯ 🏠	53	71	15	139
59	Gemma 4 12B Q8_0 ↯ 🏠	47	73	22	142
60	Gemma 4 E4B Q4_K_M ↯ 🏠	47	106	13	166
61	Qwen3.6 35B A3B Q4_K_M 💡 on* 🏠	44	79	11	134
62	Gemma 4 12B QAT-Q4_0 ↯ 🏠	41	112	9	162
63	Gemma 4 E4B Q8_0 ↯ 🏠	40	87	18	145
64	Nemotron 3 Nano Omni 30B A3B Q8_0 💡 on* 🏠	39	100	19	158
65	Granite 4.2 30B Q6_K 💡 full 🏠	39	106	6	151
66	Gemma 4 E4B Q6_K ↯ 🏠	39	126	7	172
67	Granite 4.2 30B Q4_K_M 💡 full 🏠	39	172	8	219
68	Mistral Medium 3.5 128B Q4_K_M ↯ 🏠	38	96	17	151
69	Gemma 4 E4B Q6_K 💡 on* 🏠	37	100	25	162
70	Nemotron 3 Super 120B A12B Q4_K_M 💡 full 🏠	36	102	8	146
71	Gemma 4 E4B Q8_0 💡 on* 🏠	36	119	4	159
72	Gemma 4 12B Q6_K ↯ 🏠	32	131	7	170
73	Gemma 4 E2B Q8_0 ↯ 🏠	30	131	4	165
74	Gemma 4 E2B Q8_0 💡 on* 🏠	25	104	10	139
75	Nemotron 3 Nano Omni 30B A3B Q4_K_M 💡 on* 🏠	24	112	8	144
76	Gemma 4 E2B Q6_K 💡 on* 🏠	23	133	12	168
77	Mistral Medium 3.5 💡 high	23	133	4	160
78	Gemma 4 E4B Q4_K_M 💡 on* 🏠	20	136	12	168
79	Gemma 4 E2B Q4_K_M ↯ 🏠	20	142	2	164
80	Gemma 4 E2B Q4_K_M 💡 on* 🏠	13	114	9	136
81	Gemma 4 E2B Q6_K ↯ 🏠	10	158	1	169

The AIs have had their say. Your turn.

An AI judge can read thousands of duels without coffee or a lunch break. It still cannot tell us what makes you laugh, moves you, or keeps you reading. Your blind vote feeds a separate human ranking and helps reveal where human taste and machine judgment part ways.

Judge two anonymous texts