Traces

Every run: the agent's full trajectory, its submissions over the 20-hour budget, and the final score.

Fitness-Recap Video in Remotion

task page →50 runs · indexed 2026-09-02

#	Model	Reward	Cost	In tok	Out tok	Time	Steps	Submits
1	GPT-5.6	0.2152	$950.51	938.4M	1.0M	19h 3m	2343	45
2	Claude Fable 5.1	0.2055	$157.69	293.5M	491k	6h 51m	665	8
3	Claude Fable 5.1	0.2018	$157.38	293.1M	476k	6h 7m	644	7
4	GPT-5.6	0.1930	$656.93	636.3M	650k	19h 12m	1478	11
5	Claude Fable 5.1	0.1927	$120.12	223.7M	414k	5h 14m	587	7
6	GPT-5.6	0.1913	$298.28	288.7M	261k	7h 23m	651	12
7	GPT-5.6	0.1895	$283.99	275.5M	274k	5h 43m	687	6
8	GLM-5.3	0.1894	$78.50	212.7M	442k	18h 20m	530	6
9	GLM-5.3	0.1828	$113.14	407.8M	483k	19h 2m	804	9
10	Kimi K3	0.1767	$191.75	571.0M	982k	19h 53m	1356	3
11	Claude Fable 5.1	0.1730	$560.60	1048.2M	1.5M	18h 2m	2274	26
12	Kimi K3	0.1723	$171.83	504.1M	1.1M	19h 44m	1234	8
13	Kimi K3	0.1716	$266.42	802.2M	1.3M	20h 1m	1618	1
14	DeepSeek V4 Flash Vision Exp	0.1621	$10.73	658.6M	994k	19h 13m	1651	18
15	Qwen3.8-Max	0.1474	$77.77	271.1M	994k	13h 7m	800	4
16	GLM-5.3	0.1426	$118.44	380.0M	784k	18h 55m	986	18
17	GPT-5.6	0.1374	$190.12	189.9M	230k	5h 21m	589	2
18	GLM-5.3	0.1053	$110.52	388.9M	908k	15h 36m	961	9
19	GLM-5.3	0.0837	$90.93	307.2M	485k	16h 10m	636	8
20	DeepSeek V4 Flash Vision Exp	0.0738	$6.53	405.2M	565k	19h 26m	994	10
21	Qwen3.8-Max	0.0733	$163.48	458.4M	1.7M	19h 54m	1249	1
22	Claude Fable 5.1	0.0694	$335.29	627.1M	960k	15h 45m	1456	19
23	Kimi K3	0.0657	$243.01	727.7M	1.1M	20h 1m	1633	1
24	Kimi K3	0.0536	$202.63	587.9M	1.0M	19h 57m	1479	12
25	DeepSeek V4 Flash Vision Exp	0.0483	$6.64	411.0M	525k	16h 0m	983	21
26	Grok 4.6	0.0338	$1762.94	1204.0M	2.8M	20h 0m	4175	289
27	DeepSeek V4 Flash Vision Exp	0.0324	$12.46	767.8M	1.1M	18h 4m	1789	24
28	Grok 4.6	0.0279	$646.10	557.5M	1.5M	19h 48m	2043	149
29	Grok 4.6	0.0245	$555.19	490.3M	1.5M	19h 56m	2020	90
30	Grok 4.6	0.0242	$1149.83	792.5M	1.8M	19h 58m	2836	250
31	Grok 4.6	0.0216	$746.94	531.5M	1.4M	19h 55m	1941	163
32	Gemini 3.7 Flash	0.0129	$13.97	114.7M	211k	4h 23m	382	2
33	Qwen3.8-Max	0.0127	$196.38	482.4M	1.8M	19h 51m	1048	1
34	DeepSeek V4 Flash Vision Exp	0.0121	$4.01	242.4M	395k	12h 19m	777	16
35	Gemini 3.7 Flash	0.0091	$2.94	21.2M	86k	1h 20m	169	2
36	Gemini 3.7 Flash	0.0089	$10.44	67.7M	164k	2h 48m	328	4
37	Gemini 3.7 Flash	0.0080	$7.68	61.7M	148k	2h 7m	300	1
38	Muse Spark 1.2	0.0079	$12.46	62.2M	180k	3h 28m	270	3
39	Qwen3.8-Max	0.0062	$52.83	157.6M	675k	9h 28m	578	4
40	Muse Spark 1.2	0.0060	$91.70	184.1M	1.3M	8h 42m	509	3
41	Gemini 3.7 Flash	0.0060	$4.67	30.1M	112k	2h 7m	222	1
42	Muse Spark 1.2	0.0032	$14.71	29.2M	224k	3h 30m	183	2
43	Muse Spark 1.2	0.0031	$37.85	75.3M	303k	4h 52m	285	4
44	Inkling	0.0006	$14.05	14.6M	38k	1h 0m	183	1
45	Inkling	0.0001	$7.96	8.3M	26k	31m	141	2
46	Inkling	0.0000	$13.72	14.2M	49k	59m	205	1
47	Inkling	0.0000	$10.65	11.1M	29k	1h 18m	172	2
48	Inkling	0.0000	$11.15	11.6M	44k	40m	138	1
49	Qwen3.8-Max	0.0000†	$11.91	31.2M	238k	4h 2m	292	2
50	Muse Spark 1.2	0.0000†	$12.66	38.1M	209k	4h 41m	190	1

† score adjusted by manual review; the trace's Judge tab carries the measured score and the rationale.