Name: DuelLab GameBench 2 public release gb2-20260824-e05a1ff89a26
Creator: DuelLab
License: https://creativecommons.org/licenses/by/4.0/

Game profile

Chess

Chess is an 8x8 tactical piece-movement game with asymmetric pieces, castling, en passant, promotion, check, checkmate, and automatic draw rules.

Scores here compare player programs using DuelLab's encoded rules for Chess; they are not a broad measure of human play strength.

By model and setting

GB2-02: Chess by model and setting

Player programs are ranked by relative per-game score (0–100). Raw rating is shown as an advanced metric, alongside wins, draws, losses, and a per-game uncertainty index.

Reasoning level: Balanced Game: GB2-02: Chess

GB2-02: Chess - Balanced filter
Rank	Model	Score	Contributing programs	Raw Elo	W / D / L	Uncertainty
1	Claude Fable 5	64.8	1	1673.8	31 / 27 / 6	18.3
2	Claude Opus 5	63.4	2	1668.3	60 / 54 / 10	19.2
3	GPT-5.6 Sol	54.5	1	1626.3	27 / 40 / 1	16.4
4	GPT-5.5	40.3	1	1565.5	17 / 35 / 8	20.3
5	GPT-5.6 Luna	30.1	1	1517.6	1 / 63 / 2	17.3
6	Muse Spark 1.2	28.9	1	1512.0	11 / 50 / 6	16.9
7	Claude Sonnet 5	28.9	1	1513.7	4 / 55 / 3	19.2
8	Grok 4.5	28.2	1	1503.4	10 / 61 / 13	10.5
9	Gemini 3.5 Flash	26.7	1	1499.3	10 / 59 / 7	13.2
10	GPT-5.4	25.9	1	1495.2	5 / 68 / 4	12.9
11	LongCat 2.0	24.1	1	1480.6	2 / 101 / 3	4.6
12	Kimi K2.7 Code	23.4	1	1477.9	3 / 83 / 18	5.0
13	Step 3.7 Flash	23.0	1	1483.2	1 / 67 / 6	14.0
14	Gemini 3.5 Flash Lite	22.7	1	1474.5	0 / 103 / 2	4.8
15	MiMo-V2.5	22.7	3	1481.8	6 / 195 / 19	14.0
16	Qwen3.7 Plus	22.6	1	1481.2	1 / 69 / 4	14.0
17	Claude Opus 4.5	22.4	1	1469.0	2 / 118 / 10	0.0
18	GPT-5.6 Terra	21.9	1	1478.2	0 / 67 / 7	14.0
19	GPT-5.4 Mini	21.9	1	1475.9	3 / 77 / 2	11.1
20	Grok Build 0.1	21.9	1	1477.0	3 / 73 / 2	12.5
21	GPT-OSS 120B	21.7	1	1477.3	4 / 65 / 5	14.0
22	GPT-5.4 Nano	21.5	1	1477.9	0 / 68 / 2	15.6
23	Laguna S 2.1	21.4	2	1466.6	0 / 223 / 9	2.5
24	Kimi K3	20.3	1	1464.1	7 / 78 / 17	5.5
25	Qwen3.7 Max	19.7	1	1465.4	3 / 72 / 9	10.5
26	GLM-5.2	19.2	1	1473.9	0 / 53 / 1	23.7
27	Nemotron 3 Ultra 550B A55B	17.4	1	1457.4	1 / 71 / 4	13.2
28	DeepSeek V4 Flash	17.0	2	1461.9	1 / 100 / 15	21.4
29	Gemini 3.6 Flash	16.2	1	1450.8	1 / 74 / 5	11.8
30	Claude Opus 4.8	16.0	1	1452.9	1 / 64 / 5	15.6
31	Inkling Small	14.2	1	1484.9	0 / 20 / 0	65.4
32	Ling 3.0 Flash	13.0	2	1458.0	2 / 95 / 21	38.5
33	Qwen3.8 Max	11.8	1	1474.2	0 / 18 / 2	65.4
34	Ox Alpha	10.0	3	1482.2	7 / 24 / 11	85.3
35	Minimax M3	9.5	1	1422.7	1 / 59 / 14	14.0

Underlying results

Individual player programs

Each row is one accepted player program for GB2-02: Chess. A model can appear more than once when it produced more than one program. Score is the normalized 0–100 result; Elo and W / D / L show the underlying play record. Confidence and evidence can be blank when too few compatible matches are available. Draws at the move limit count; one-sided model faults are listed separately.

Search model Status Sort Showing 43 player programs.

Player ID	Model	Status	Score	Elo	Confidence	W / D / L	Rated / attempted	Opponents	Faults / players	Generation
E-cd3608270e8c00fd	Claude Fable 5	Rated	64.8	1674	—	31 / 27 / 6	64 / 66	30	2 / 1	First try · 2026-07-02
E-03472a04778e9928	Claude Opus 4.5	Rated	22.4	1469	—	2 / 118 / 10	130 / 130	50	0 / 1	First try · 2026-07-21
E-3cbd11fb314c75e1	Claude Opus 4.5	Waiting	—	—	—	0 / 0 / 0	0 / 0	0	0 / 1	First try · 2026-07-02
E-ed71c4ecf219b609	Claude Opus 4.8	Rated	16.0	1453	—	1 / 64 / 5	70 / 70	35	0 / 1	First try · 2026-07-02
E-50b401cafe71e710	Claude Opus 5	Rated	62.0	1662	—	30 / 26 / 6	62 / 62	31	0 / 1	First try · 2026-07-25
E-8a5bf2e23121edf6	Claude Opus 5	Rated	64.8	1674	—	30 / 28 / 4	62 / 62	31	0 / 1	First try · 2026-07-25
E-874db7679e1f7332	Claude Sonnet 5	Rated	28.9	1514	—	4 / 55 / 3	62 / 62	30	0 / 1	First try · 2026-08-09
E-08e2baea96c43eb4	DeepSeek V4 Flash	Rated	13.5	1447	—	0 / 50 / 6	56 / 56	28	0 / 1	First try · 2026-07-02
E-1e168b9e73c26073	DeepSeek V4 Flash	Rated	20.4	1476	—	1 / 50 / 9	60 / 60	30	0 / 1	Repaired · 2026-07-01
E-ef7a835bdb46a90e	Gemini 3.5 Flash	Rated	26.7	1499	—	10 / 59 / 7	76 / 76	38	0 / 1	Repaired · 2026-07-02
E-dbbc5835674b5f2c	Gemini 3.5 Flash Lite	Rated	22.7	1475	—	0 / 103 / 2	105 / 105	40	0 / 1	First try · 2026-07-24
E-62275b0c39dda32a	Gemini 3.6 Flash	Rated	16.2	1451	—	1 / 74 / 5	80 / 80	40	0 / 1	First try · 2026-07-24
E-67423a89b95a1aba	GPT-5.4	Rated	25.9	1495	—	5 / 68 / 4	77 / 77	34	0 / 1	First try · 2026-08-09
E-6724998b59176607	GPT-5.4 Mini	Rated	21.9	1476	—	3 / 77 / 2	82 / 82	39	0 / 1	First try · 2026-07-02
E-d96eaa4f1d20c47b	GPT-5.4 Nano	Rated	21.5	1478	—	0 / 68 / 2	70 / 70	33	0 / 1	Repaired · 2026-07-02
E-8460841e3919c568	GPT-5.5	Rated	40.3	1565	—	17 / 35 / 8	60 / 60	30	0 / 1	First try · 2026-07-02
E-2f3ae5ed2aa83217	GPT-5.6 Luna	Rated	30.1	1518	—	1 / 63 / 2	66 / 66	30	0 / 1	First try · 2026-08-09
E-1d0646e1538c1b7a	GPT-5.6 Sol	Rated	54.5	1626	—	27 / 40 / 1	68 / 68	30	0 / 1	First try · 2026-08-09
E-82ad98d8d570daab	GPT-5.6 Terra	Rated	21.9	1478	—	0 / 67 / 7	74 / 74	36	0 / 1	Repaired · 2026-07-10
E-129026c5e53b35ad	Ling 3.0 Flash	Limited data	6.0	1453	—	0 / 15 / 3	18 / 18	9	0 / 1	Repaired · 2026-07-24
E-a00d8e3eb7dd07cf	LongCat 2.0	Rated	24.1	1481	—	2 / 101 / 3	106 / 106	41	0 / 1	Repaired · 2026-07-24
E-2b12f5618349c056	Muse Spark 1.2	Rated	28.9	1512	—	11 / 50 / 6	67 / 67	31	0 / 1	First try · 2026-08-13
E-cad87bcb23d33fef	Minimax M3	Rated	9.5	1423	—	1 / 59 / 14	74 / 74	33	0 / 1	Repaired · 2026-07-02
E-85c8a9372a607001	Kimi K2.7 Code	Rated	23.4	1478	—	3 / 83 / 18	104 / 104	49	0 / 1	Repaired · 2026-07-02
E-0c4144684fc36307	Kimi K3	Rated	20.3	1464	—	7 / 78 / 17	102 / 102	38	0 / 1	First try · 2026-07-24
E-bd70ebbd81fdbef0	Nemotron 3 Ultra 550B A55B	Rated	17.4	1457	—	1 / 71 / 4	76 / 76	37	0 / 1	Repaired · 2026-07-02
E-8f0c6569aa132f8e	GPT-OSS 120B	Rated	21.7	1477	—	4 / 65 / 5	74 / 74	36	0 / 1	First try · 2026-07-02
E-7164ee63b9dd911a	Laguna S 2.1	Rated	22.1	1470	—	0 / 112 / 4	116 / 118	41	0 / 1	First try · 2026-07-24
E-ba56689188a301b2	Laguna S 2.1	Rated	20.6	1463	—	0 / 111 / 5	116 / 116	43	0 / 1	Repaired · 2026-07-24
E-95265340f8562bfd	Qwen3.7 Max	Rated	19.7	1465	—	3 / 72 / 9	84 / 84	41	0 / 1	Repaired · 2026-07-02
E-57293d5b71f14882	Qwen3.7 Plus	Rated	22.6	1481	—	1 / 69 / 4	74 / 74	34	0 / 1	Repaired · 2026-07-02
E-7ca9241c6a1cf54f	Qwen3.8 Max	Limited data	11.8	1474	—	0 / 18 / 2	20 / 20	10	0 / 1	First try · 2026-08-08
E-427d180a6aa40aa2	Ox Alpha	Rated	18.7	1521	—	4 / 8 / 2	14 / 14	7	0 / 1	First try · 2026-08-22
E-8ad03916301062ed	Ox Alpha	Rated	11.4	1488	—	2 / 9 / 3	14 / 14	7	0 / 1	First try · 2026-08-22
E-b4db3ddedfd59608	Ox Alpha	Rated	0.0	1437	—	1 / 7 / 6	14 / 14	7	0 / 1	First try · 2026-08-22
E-0ccea44addccfff0	Step 3.7 Flash	Rated	23.0	1483	—	1 / 67 / 6	74 / 74	35	0 / 1	Repaired · 2026-07-02
E-647caf2136958821	Inkling Small	Limited data	14.2	1485	—	0 / 20 / 0	20 / 20	11	0 / 1	Repaired · 2026-08-09
E-da5ea0258e71b6ab	Grok 4.5	Rated	28.2	1503	—	10 / 61 / 13	84 / 84	42	0 / 1	First try · 2026-07-24
E-99592fa0331c6679	Grok Build 0.1	Rated	21.9	1477	—	3 / 73 / 2	78 / 78	37	0 / 1	Repaired · 2026-07-02
E-5b213ae773e821cc	MiMo-V2.5	Rated	25.2	1492	—	3 / 63 / 10	76 / 76	34	0 / 1	Repaired · 2026-07-02
E-a5f372b0e5962ee5	MiMo-V2.5	Rated	17.9	1462	—	0 / 64 / 4	68 / 68	32	0 / 1	Repaired · 2026-07-01
E-c37f25e181320efb	MiMo-V2.5	Rated	22.6	1481	—	3 / 68 / 5	76 / 76	36	0 / 1	Repaired · 2026-07-02
E-32435b01ac08ed41	GLM-5.2	Rated	19.2	1474	—	0 / 53 / 1	54 / 54	27	0 / 1	Repaired · 2026-07-11