Datasets:

open-agent-leaderboard
/

results

agent large_stringclasses 5 values	agent_name large_stringclasses 5 values	average_action_count float64 0 77.4	average_agent_cost float64 0 22.8	average_benchmark_cost float64 0 0.13	average_invalid_action_count float64 0 39.1	average_invalid_action_percent float64 0 75.7	average_score float64 0 0.88	average_steps float64 0 77.4	benchmark large_stringclasses 6 values	benchmark_name large_stringclasses 4 values	benchmark_score float64 0 0.88	completed_sessions float64 0 100	incomplete_sessions float64 0 100	model large_stringclasses 5 values	model_name large_stringclasses 5 values	percent_error float64 0 1	percent_finished float64 0 1	percent_finished_successful float64 0 1	percent_finished_unsuccessful float64 0 1	percent_successful float64 0 1	percent_unfinished float64 0 0.54	planned_sessions int64 50 100	subset_name large_stringclasses 4 values	successful_sessions int64 0 100	total_agent_cost float64 0 2.28k	total_benchmark_cost float64 0 12.8	total_run_cost float64 0 2.28k	total_sessions int64 50 100
claude_code	Claude Code CLI	18.94	0.502607	0	7.87	41.55227	0.44676	18.94	appworld_test_normal	AppWorld	0.03	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.46	0.03	0.43	0.03	0.54	100	test_normal	3	50.260664	0	50.260664	100
claude_code	Claude Code CLI	20.82	0.106301	0.001696	0	0	0.48	20.82	browsecompplus	BrowseCompPlus	0.48	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.98	0.48	0.5	0.48	0.02	100	null	48	10.630119	0.169576	10.799695	100
claude_code	Claude Code CLI	54.16	0.201441	0	0	0	0.64	54.16	swebench	SWE-bench	0.64	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.89	0.89	0	0.89	0.11	100	null	89	20.144125	0	20.144125	100
claude_code	Claude Code CLI	12.06	0.028656	0.011808	0	0	0.28	12.06	tau2_airline	Tau Bench 2	0.28	50	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.98	0.28	0.7	0.28	0.02	50	airline	14	1.432799	0.590388	2.023187	50
claude_code	Claude Code CLI	12.51	0.043596	0.012948	0	0	0.65	12.51	tau2_retail	Tau Bench 2	0.65	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.99	0.65	0.34	0.65	0.01	100	retail	65	4.359629	1.294778	5.654407	100
claude_code	Claude Code CLI	20.03	0.072257	0.098636	0	0	0.61	20.03	tau2_telecom	Tau Bench 2	0.61	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.97	0.61	0.36	0.61	0.03	100	telecom	61	7.22567	9.863552	17.089222	100
claude_code	Claude Code CLI	36.62	0.893032	0	13.48	36.810486	0.55621	36.62	appworld_test_normal	AppWorld	0.08	100	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	0.78	0.08	0.7	0.08	0.22	100	test_normal	8	89.303152	0	89.303152	100
claude_code	Claude Code CLI	25.02	0.732051	0.001531	0	0	0.56	25.02	browsecompplus	BrowseCompPlus	0.56	100	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	0.86	0.56	0.3	0.56	0.14	100	null	56	73.205077	0.15312	73.358197	100
claude_code	Claude Code CLI	38.484848	0.782263	0	0	0	0.515152	38.484848	swebench	SWE-bench	0.520408	98	2	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0.010101	0.747475	0.747475	0	0.747475	0.242424	100	null	74	77.444029	0	77.444029	99
claude_code	Claude Code CLI	2.58	0.032655	0.002739	0	0	0.12	2.58	tau2_airline	Tau Bench 2	0.12	50	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	0.9	0.12	0.78	0.12	0.1	50	airline	6	1.632725	0.136938	1.769663	50
claude_code	Claude Code CLI	0.56	0.01857	0.001576	0	0	0.03	0.56	tau2_retail	Tau Bench 2	0.03	100	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	1	0.03	0.97	0.03	0	100	retail	3	1.857011	0.157608	2.014619	100
claude_code	Claude Code CLI	0	0.005501	0.005097	0	0	0	0	tau2_telecom	Tau Bench 2	0	100	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	1	0	1	0	0	100	telecom	0	0.550054	0.509724	1.059778	100
claude_code	Claude Code CLI	0	0	0	0	0	0	0	appworld_test_normal	AppWorld	0	0	100	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	1	0	0	0	0	0	100	test_normal	0	0	0	0	100
claude_code	Claude Code CLI	0	0.429381	0.001692	0	0	0.43	8.97	browsecompplus	BrowseCompPlus	0.43	100	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.43	0.57	0.43	0	100	null	43	42.93807	0.169218	43.107288	100
claude_code	Claude Code CLI	1.63	0.939755	0	0	0	0.58	23.99	swebench	SWE-bench	0.58	100	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	1	0	1	0	100	null	100	93.975473	0	93.975473	100
claude_code	Claude Code CLI	4.12	0.213383	0.011296	0	0	0.48	10.18	tau2_airline	Tau Bench 2	0.48	50	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.48	0.52	0.48	0	50	airline	24	10.669148	0.564818	11.233966	50
claude_code	Claude Code CLI	6.68	0.120064	0.006263	0	0	0.51	9.92	tau2_retail	Tau Bench 2	0.64	100	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	0.98	0.64	0.34	0.64	0.02	100	retail	64	12.006437	0.626275	12.632711	100
claude_code	Claude Code CLI	6.57	0.098015	0.053513	0	0	0.55	9.36	tau2_telecom	Tau Bench 2	0.55	100	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.55	0.45	0.55	0	100	telecom	55	9.801496	5.351322	15.152818	100
claude_code	Claude Code CLI	49.69	13.083792	0	18.17	36.566714	0.77417	49.69	appworld_test_normal	AppWorld	0.66	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	0.74	0.66	0.08	0.66	0.26	100	test_normal	66	1,308.379215	0	1,308.379215	100
claude_code	Claude Code CLI	14.333333	11.658796	0.001639	0	0	0.529412	31.039216	browsecompplus	BrowseCompPlus	0.529412	0	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	0.843137	0.529412	0.313725	0.529412	0.156863	51	null	27	594.598615	0.083574	594.682189	51
claude_code	Claude Code CLI	0	5.604361	0	0	0	0.742268	31.762887	swebench	SWE-bench	0.742268	97	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	1	0	1	0	97	null	97	543.623015	0	543.623015	97
claude_code	Claude Code CLI	11.5	1.299355	0.01386	0	0	0.66	11.5	tau2_airline	Tau Bench 2	0.66	50	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.66	0.34	0.66	0	50	airline	33	64.967745	0.692992	65.660737	50
claude_code	Claude Code CLI	12.54	1.598523	0.012835	0	0	0.83	12.54	tau2_retail	Tau Bench 2	0.83	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.83	0.17	0.83	0	100	retail	83	159.852325	1.283548	161.135873	100
claude_code	Claude Code CLI	18.71	2.448865	0.110861	0	0	0.76	18.71	tau2_telecom	Tau Bench 2	0.76	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.76	0.24	0.76	0	100	telecom	76	244.88646	11.086118	255.972578	100
claude_code	Claude Code CLI	38.01	3.105508	0	17.26	45.409103	0.72641	38.01	appworld_test_normal	AppWorld	0.36	100	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	0.86	0.36	0.5	0.36	0.14	100	test_normal	36	310.550762	0	310.550762	100
claude_code	Claude Code CLI	0	2.845294	0.001463	0	0	0.51	22.88	browsecompplus	BrowseCompPlus	0.51	0	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	0.7	0.51	0.19	0.51	0.3	100	null	51	284.529442	0.146346	284.675788	100
claude_code	Claude Code CLI	0	3.679669	0	0	0	0.67	43.72	swebench	SWE-bench	0.67	100	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	0.99	0.01	0.99	0	100	null	99	367.966876	0	367.966876	100
claude_code	Claude Code CLI	12.62	0.339443	0.009578	0	0	0.7	12.62	tau2_airline	Tau Bench 2	0.7	50	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	0.7	0.3	0.7	0	50	airline	35	16.97213	0.478889	17.451019	50
claude_code	Claude Code CLI	11.18	0.185258	0.008511	0	0	0.71	11.18	tau2_retail	Tau Bench 2	0.71	100	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	0.71	0.29	0.71	0	100	retail	71	18.525812	0.851069	19.376881	100
claude_code	Claude Code CLI	7.16	0.210642	0.044182	0	0	0.71	9.9	tau2_telecom	Tau Bench 2	0.71	100	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	0.71	0.29	0.71	0	100	telecom	71	21.064224	4.41815	25.482374	100
openai_solo	OpenAI Solo	15.63	0.305088	0	6.3	40.307102	0.29724	15.63	appworld_test_normal	AppWorld	0.06	48	52	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0.52	0.37	0.05	0.32	0.06	0.11	100	test_normal	6	30.508816	0	30.508816	100
openai_solo	OpenAI Solo	13.13	0.095325	0.001036	0	0	0.3	13.13	browsecompplus	BrowseCompPlus	0.3	62	38	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0.38	0.62	0.3	0.32	0.3	0	100	null	30	9.532471	0.103588	9.636059	100
openai_solo	OpenAI Solo	41.37	0.112377	0	0	0	0.28	41.37	swebench	SWE-bench	0.736842	38	62	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0.62	0.38	0.38	0	0.38	0	100	null	38	11.237704	0	11.237704	100
openai_solo	OpenAI Solo	6.72	0.016995	0.005967	0	0	0.2	6.72	tau2_airline	Tau Bench 2	0.2	50	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.94	0.2	0.74	0.2	0.06	50	airline	10	0.849731	0.298334	1.148065	50
openai_solo	OpenAI Solo	4.82	0.011356	0.005682	0	0	0.19	4.82	tau2_retail	Tau Bench 2	0.19	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.97	0.19	0.78	0.19	0.03	100	retail	19	1.135619	0.568166	1.703785	100
openai_solo	OpenAI Solo	7	0.023369	0.032554	0	0	0.18	7	tau2_telecom	Tau Bench 2	0.18	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.97	0.18	0.79	0.18	0.03	100	telecom	18	2.336938	3.255376	5.592314	100
openai_solo	OpenAI Solo	19.59	0.302422	0	7.6	38.795304	0.34422	19.59	appworld_test_normal	AppWorld	0.08	52	48	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0.48	0.48	0.08	0.4	0.08	0.04	100	test_normal	8	30.24223	0	30.24223	100
openai_solo	OpenAI Solo	17.61	0.251447	0.000974	0	0	0.35	17.61	browsecompplus	BrowseCompPlus	0.35	62	38	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0.38	0.53	0.35	0.18	0.35	0.09	100	null	35	25.144735	0.097406	25.242141	100
openai_solo	OpenAI Solo	38.74	0.334183	0	0	0	0.55	38.74	swebench	SWE-bench	0.56701	97	3	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0.03	0.78	0.78	0	0.78	0.19	100	null	78	33.418318	0	33.418318	100
openai_solo	OpenAI Solo	0.5	0.004579	0.001447	0	0	0	0.5	tau2_airline	Tau Bench 2	0	50	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	0.88	0	0.88	0	0.12	50	airline	0	0.228964	0.072342	0.301306	50
openai_solo	OpenAI Solo	0.41	0.004281	0.001557	0	0	0.01	0.41	tau2_retail	Tau Bench 2	0.01	100	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	0.96	0.01	0.95	0.01	0.04	100	retail	1	0.428117	0.155726	0.583843	100
openai_solo	OpenAI Solo	0	0.00464	0.005097	0	0	0	0	tau2_telecom	Tau Bench 2	0	100	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	1	0	1	0	0	100	telecom	0	0.463963	0.509724	0.973687	100
openai_solo	OpenAI Solo	0	0	0	0	0	0	0	appworld_test_normal	AppWorld	0	0	100	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	1	0	0	0	0	0	100	test_normal	0	0	0	0	100
openai_solo	OpenAI Solo	0	0.380437	0.001675	0	0	0.48	14.27	browsecompplus	BrowseCompPlus	0.48	100	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.48	0.52	0.48	0	100	null	48	38.043726	0.167464	38.21119	100
openai_solo	OpenAI Solo	0	0.259018	0	0	0	0.545455	20.444444	swebench	SWE-bench	0.545455	99	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.989899	0.010101	0.989899	0	99	null	98	25.642766	0	25.642766	99
openai_solo	OpenAI Solo	11.4	0.107118	0.008216	0	0	0.5	11.4	tau2_airline	Tau Bench 2	0.5	50	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.5	0.5	0.5	0	50	airline	25	5.355903	0.410809	5.766712	50
openai_solo	OpenAI Solo	9.55	0.107783	0.007619	0	0	0.53	9.55	tau2_retail	Tau Bench 2	0.53	100	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	0.99	0.53	0.46	0.53	0.01	100	retail	53	10.77833	0.761935	11.540265	100
openai_solo	OpenAI Solo	9.92	0.147175	0.04162	0	0	0.53	9.92	tau2_telecom	Tau Bench 2	0.53	100	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.53	0.47	0.53	0	100	telecom	53	14.717546	4.162018	18.879564	100
openai_solo	OpenAI Solo	47.65	22.764768	0	18.16	38.111228	0.79381	47.65	appworld_test_normal	AppWorld	0.68	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	0.77	0.68	0.09	0.68	0.23	100	test_normal	68	2,276.47676	0	2,276.47676	100
openai_solo	OpenAI Solo	0	7.592457	0.001918	0	0	0.61	27.18	browsecompplus	BrowseCompPlus	0.61	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.61	0.39	0.61	0	100	null	61	759.245735	0.191776	759.437511	100
openai_solo	OpenAI Solo	7.433735	2.961261	0	0	0	0.807229	34.096386	swebench	SWE-bench	0.807229	83	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	1	0	1	0	83	null	83	245.78468	0	245.78468	83
openai_solo	OpenAI Solo	12.22	0.716584	0.014479	0	0	0.74	12.22	tau2_airline	Tau Bench 2	0.74	50	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.74	0.26	0.74	0	50	airline	37	35.82921	0.723941	36.553151	50
openai_solo	OpenAI Solo	12.54	0.549581	0.012196	0	0	0.85	12.54	tau2_retail	Tau Bench 2	0.85	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.85	0.15	0.85	0	100	retail	85	54.95812	1.219595	56.177715	100
openai_solo	OpenAI Solo	17.15	1.24594	0.122447	0	0	0.84	17.15	tau2_telecom	Tau Bench 2	0.84	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.84	0.16	0.84	0	100	telecom	84	124.59395	12.244745	136.838695	100
openai_solo	OpenAI Solo	33.49	8.695528	0	13.52	40.37026	0.81766	33.49	appworld_test_normal	AppWorld	0.57	98	2	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0.02	0.98	0.57	0.41	0.57	0	100	test_normal	57	869.552842	0	869.552842	100
openai_solo	OpenAI Solo	0	0.643307	0.001068	0	0	0.333333	8.454545	browsecompplus	BrowseCompPlus	0.333333	0	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0.020202	0.606061	0.333333	0.272727	0.333333	0.373737	99	null	33	63.687404	0.105778	63.793182	99
openai_solo	OpenAI Solo	6.531915	1.579178	0	0	0	0.723404	32.361702	swebench	SWE-bench	0.723404	94	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	1	0	1	0	94	null	94	148.44273	0	148.44273	94
openai_solo	OpenAI Solo	10.9	0.213802	0.009739	0	0	0.62	10.9	tau2_airline	Tau Bench 2	0.62	50	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	0.62	0.38	0.62	0	50	airline	31	10.690106	0.48695	11.177056	50
openai_solo	OpenAI Solo	10.62	0.266485	0.008298	0	0	0.73	10.62	tau2_retail	Tau Bench 2	0.73	100	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	0.73	0.27	0.73	0	100	retail	73	26.6485	0.829764	27.478264	100
openai_solo	OpenAI Solo	10.82	0.535132	0.047736	0	0	0.79	10.82	tau2_telecom	Tau Bench 2	0.79	89	11	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0.11	0.89	0.79	0.1	0.79	0	100	telecom	79	53.513152	4.773591	58.286743	100
smolagents_code	SmolAgents Code	54.69	0.242061	0	25.24	46.151033	0.61274	54.69	appworld_test_normal	AppWorld	0.13	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.81	0.13	0.68	0.13	0.19	100	test_normal	13	24.20606	0	24.20606	100
smolagents_code	SmolAgents Code	12.01	0.077684	0.001468	0	0	0.21	12.01	browsecompplus	BrowseCompPlus	0.21	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	1	0.21	0.79	0.21	0	100	null	21	7.768438	0.146832	7.91527	100
smolagents_code	SmolAgents Code	69.43	0.551922	0	0	0	0.56	69.43	swebench	SWE-bench	0.56	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	0.98	0.97	0.01	0.97	0.02	100	null	97	55.192238	0	55.192238	100
smolagents_code	SmolAgents Code	18.94	0.055421	0.011782	0	0	0.6	18.94	tau2_airline	Tau Bench 2	0.6	50	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	1	0.6	0.4	0.6	0	50	airline	30	2.771043	0.58908	3.360123	50
smolagents_code	SmolAgents Code	14.06	0.051291	0.010719	0	0	0.77	14.06	tau2_retail	Tau Bench 2	0.77	98	2	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0.02	0.98	0.77	0.21	0.77	0	100	retail	77	5.12911	1.071856	6.200966	100
smolagents_code	SmolAgents Code	18.3	0.068068	0.07849	0	0	0.84	18.3	tau2_telecom	Tau Bench 2	0.84	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	1	0.84	0.16	0.84	0	100	telecom	84	6.806757	7.849034	14.655791	100
smolagents_code	SmolAgents Code	50.63	0.674029	0	9.18	18.131543	0.61839	50.63	appworld_test_normal	AppWorld	0.11	100	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	0.91	0.11	0.8	0.11	0.09	100	test_normal	11	67.402861	0	67.402861	100
smolagents_code	SmolAgents Code	29.79	0.494878	0.0015	0	0	0.33	29.79	browsecompplus	BrowseCompPlus	0.33	100	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	0.95	0.33	0.62	0.33	0.05	100	null	33	49.48785	0.150022	49.637872	100
smolagents_code	SmolAgents Code	65.913978	1.374179	0	0	0	0.569892	65.913978	swebench	SWE-bench	0.576087	92	8	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0.010753	0.870968	0.870968	0	0.870968	0.11828	100	null	81	127.798619	0	127.798619	93
smolagents_code	SmolAgents Code	13	0.099287	0.012204	0	0	0.56	13	tau2_airline	Tau Bench 2	0.56	50	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	1	0.56	0.44	0.56	0	50	airline	28	4.964369	0.610222	5.574591	50
smolagents_code	SmolAgents Code	12.887755	0.067532	0.011655	0.010204	0.079177	0.72449	12.887755	tau2_retail	Tau Bench 2	0.72449	98	2	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	1	0.72449	0.27551	0.72449	0	100	retail	71	6.618142	1.142178	7.76032	98
smolagents_code	SmolAgents Code	19.141414	0.153413	0.124975	0	0	0.707071	19.141414	tau2_telecom	Tau Bench 2	0.707071	99	1	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	1	0.707071	0.292929	0.707071	0	100	telecom	70	15.187928	12.372566	27.560494	99
smolagents_code	SmolAgents Code	51.59	0.550311	0	39.05	75.692964	0.40232	51.59	appworld_test_normal	AppWorld	0.07	98	2	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0.02	0.61	0.07	0.54	0.07	0.37	100	test_normal	7	55.031142	0	55.031142	100
smolagents_code	SmolAgents Code	0	0.171568	0.001547	0	0	0.26	6.57	browsecompplus	BrowseCompPlus	0.26	0	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	0.99	0.26	0.73	0.26	0.01	100	null	26	17.156755	0.154748	17.311503	100
smolagents_code	SmolAgents Code	2.848485	0.450318	0	0	0	0.525253	19.979798	swebench	SWE-bench	0.525253	99	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	1	0	1	0	99	null	99	44.581478	0	44.581478	99
smolagents_code	SmolAgents Code	10.68	0.293272	0.012412	0	0	0.6	10.68	tau2_airline	Tau Bench 2	0.6	50	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.6	0.4	0.6	0	50	airline	30	14.663588	0.620616	15.284204	50
smolagents_code	SmolAgents Code	11.08	0.251947	0.010713	0	0	0.68	11.08	tau2_retail	Tau Bench 2	0.68	100	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.68	0.32	0.68	0	100	retail	68	25.194685	1.071263	26.265948	100
smolagents_code	SmolAgents Code	10.11	0.303553	0.049517	0	0	0.71	10.11	tau2_telecom	Tau Bench 2	0.71	100	0	openai_Azure_gpt-5.2-2025-12-11	openai/Azure/gpt-5.2-2025-12-11	0	1	0.71	0.29	0.71	0	100	telecom	71	30.355348	4.951702	35.30705	100
smolagents_code	SmolAgents Code	41.07	5.585109	0	4.8	11.687363	0.82331	41.07	appworld_test_normal	AppWorld	0.7	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	0.82	0.7	0.12	0.7	0.18	100	test_normal	70	558.51091	0	558.51091	100
smolagents_code	SmolAgents Code	0	6.303725	0.00183	0	0	0.61	24.16	browsecompplus	BrowseCompPlus	0.61	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.61	0.39	0.61	0	100	null	61	630.372505	0.183026	630.555531	100
smolagents_code	SmolAgents Code	15.77	4.852183	0	0	0	0.65	39.13	swebench	SWE-bench	0.65	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	1	0	1	0	100	null	100	485.218325	0	485.218325	100
smolagents_code	SmolAgents Code	11.88	0.780153	0.013339	0	0	0.72	11.88	tau2_airline	Tau Bench 2	0.72	50	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.72	0.28	0.72	0	50	airline	36	39.00764	0.666959	39.674599	50
smolagents_code	SmolAgents Code	11.71	0.671186	0.011224	0	0	0.78	11.71	tau2_retail	Tau Bench 2	0.78	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.78	0.22	0.78	0	100	retail	78	67.11861	1.122397	68.241007	100
smolagents_code	SmolAgents Code	13.77	1.064388	0.08183	0	0	0.58	13.77	tau2_telecom	Tau Bench 2	0.58	100	0	openai_aws_claude-opus-4-5	openai/aws/claude-opus-4-5	0	1	0.58	0.42	0.58	0	100	telecom	58	106.43879	8.183033	114.621823	100
smolagents_code	SmolAgents Code	49.13	2.542546	0	20.51	41.746387	0.526	49.13	appworld_test_normal	AppWorld	0.13	100	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	0.71	0.13	0.58	0.13	0.29	100	test_normal	13	254.254602	0	254.254602	100
smolagents_code	SmolAgents Code	0	2.388851	0.001138	0	0	0.57	29.63	browsecompplus	BrowseCompPlus	0.57	0	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	0.69	0.57	0.12	0.57	0.31	100	null	57	238.885126	0.113834	238.99896	100
smolagents_code	SmolAgents Code	0	2.209662	0	0	0	0.757576	38.10101	swebench	SWE-bench	0.757576	99	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	1	0	1	0	99	null	99	218.756522	0	218.756522	99
smolagents_code	SmolAgents Code	12.28	0.195905	0.009922	0	0	0.68	12.28	tau2_airline	Tau Bench 2	0.68	50	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	0.68	0.32	0.68	0	50	airline	34	9.795244	0.496103	10.291347	50
smolagents_code	SmolAgents Code	11.3	0.205686	0.008618	0	0	0.75	11.3	tau2_retail	Tau Bench 2	0.75	100	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	0.75	0.25	0.75	0	100	retail	75	20.56864	0.861754	21.430394	100
smolagents_code	SmolAgents Code	12.71	0.34617	0.056376	0	0	0.88	12.71	tau2_telecom	Tau Bench 2	0.88	100	0	openai_gcp_gemini-3-pro-preview	openai/gcp/gemini-3-pro-preview	0	1	0.88	0.12	0.88	0	100	telecom	88	34.61699	5.63759	40.25458	100
tool_calling	LiteLLM Tool Calling	31.7	0.833262	0	1.27	4.006309	0.5494	31.69	appworld_test_normal	AppWorld	0.09	99	1	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0.01	0.88	0.09	0.79	0.09	0.11	100	test_normal	9	83.326163	0	83.326163	100
tool_calling	LiteLLM Tool Calling	34.19	0.167917	0.00165	0.03	0.087745	0.36	34.19	browsecompplus	BrowseCompPlus	0.36	99	1	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0.01	0.98	0.36	0.62	0.36	0.01	100	null	36	16.791678	0.164996	16.956674	100
tool_calling	LiteLLM Tool Calling	77.44	0.478891	0	0.01	0.012913	0.66	77.44	swebench	SWE-bench	0.6875	96	4	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0.04	0.94	0.93	0.01	0.93	0.02	100	null	93	47.889092	0	47.889092	100
tool_calling	LiteLLM Tool Calling	15.32	0.033953	0.017188	0	0	0.56	15.28	tau2_airline	Tau Bench 2	0.56	50	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	1	0.56	0.44	0.56	0	50	airline	28	1.697626	0.859394	2.55702	50
tool_calling	LiteLLM Tool Calling	14.44	0.029599	0.018759	0	0	0.82	14.44	tau2_retail	Tau Bench 2	0.82	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	1	0.82	0.18	0.82	0	100	retail	82	2.959856	1.875946	4.835802	100
tool_calling	LiteLLM Tool Calling	18.13	0.049405	0.127713	0	0	0.71	18.11	tau2_telecom	Tau Bench 2	0.71	100	0	openai_Azure_DeepSeek-V3.2	openai/azure/DeepSeek-V3.2	0	1	0.71	0.29	0.71	0	100	telecom	71	4.940464	12.77128	17.711744	100
tool_calling	LiteLLM Tool Calling	43.03	0.892555	0	1.28	2.974669	0.60545	25.05	appworld_test_normal	AppWorld	0.09	99	1	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0.01	0.9	0.09	0.81	0.09	0.09	100	test_normal	9	89.255475	0	89.255475	100
tool_calling	LiteLLM Tool Calling	27.26	0.379113	0.001292	0.52	1.907557	0.34	26.27	browsecompplus	BrowseCompPlus	0.34	98	2	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0.02	0.74	0.34	0.4	0.34	0.24	100	null	34	37.911278	0.129216	38.040494	100
tool_calling	LiteLLM Tool Calling	48.765306	0.590102	0	0.040816	0.0837	0.571429	48.765306	swebench	SWE-bench	0.571429	98	2	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	0.877551	0.877551	0	0.877551	0.122449	100	null	86	57.830003	0	57.830003	98
tool_calling	LiteLLM Tool Calling	10.76	0.039222	0.012553	0	0	0.62	9.6	tau2_airline	Tau Bench 2	0.62	50	0	openai_Azure_Kimi-K2.5	openai/azure/Kimi-K2.5	0	1	0.62	0.38	0.62	0	50	airline	31	1.961081	0.627634	2.588715	50

End of preview. Expand in Data Studio

Open Agent Leaderboard Results

Detailed evaluation results for general-purpose AI agents across diverse real-world benchmarks — without domain-specific tuning.

Leaderboard: open-agent-leaderboard/leaderboard
Website: exgentic.ai
Paper: arXiv:2602.22953
GitHub: Exgentic/exgentic
License: CDLA-Permissive-2.0

Benchmarks

Benchmark	Task ID	Description
AppWorld	`appworld`	App-based task completion in simulated smartphone environments
BrowseComp+	`browsecomp_plus`	Web browsing and complex information retrieval
SWE-bench	`swebench`	Software engineering issue resolution on real GitHub repos
TauBench-Airline	`taubench_airline`	Customer service agent evaluation (airline domain)
TauBench-Retail	`taubench_retail`	Customer service agent evaluation (retail domain)
TauBench-Telecom	`taubench_telecom`	Customer service agent evaluation (telecom domain)

The overall score is a weighted average: each TauBench sub-task gets 1/12 weight (1/4 total for TauBench), all others get 1/4 each.

Agents Evaluated

Agent	Framework
Claude Code	claude-code
OpenAI Solo	openai-agents-python
Smolagent	smolagents
React	litellm
React + Shortlisting	litellm + exgentic

Models

Results are reported for each agent × model combination: Claude Opus 4.5, Gemini 3 Pro, GPT-5.2, DeepSeek V3.2, Kimi K2.5.

Submitting new results

This dataset is the source of truth for the Open Agent Leaderboard. To add results for a new model, agent, or benchmark:

Run evaluations using the Exgentic framework
Open a PR on this dataset adding your rows to the parquet file in data/

Each row represents one (agent, model, benchmark) combination. Required fields:

Field	Description
`agent`	Agent identifier (e.g., `claude_code`)
`agent_name`	Display name (e.g., `Claude Code CLI`)
`model`	Model identifier (e.g., `openai_Azure_DeepSeek-V3.2`)
`model_name`	Display name (e.g., `openai/azure/DeepSeek-V3.2`)
`benchmark`	Benchmark identifier (e.g., `swebench`)
`benchmark_name`	Display name (e.g., `SWE-bench`)
`benchmark_score`	Primary score (0-1)
`planned_sessions`	Number of tasks attempted
`total_sessions`	Number of sessions completed
`successful_sessions`	Number of sessions that passed

See the existing data for the full schema and examples.

Schema

Column	Type	Description
`agent` / `agent_name`	string	Agent identifier and display name
`model` / `model_name`	string	Model identifier and display name
`benchmark` / `benchmark_name`	string	Benchmark identifier and display name
`benchmark_score`	float	Primary success rate (0-1)
`average_score`	float	Average score across sessions
`average_agent_cost`	float	Average cost per task (USD)
`average_steps`	float	Average number of agent steps per task
`average_action_count`	float	Average number of actions per task
`average_invalid_action_count`	float	Average invalid actions per task
`percent_successful`	float	Fraction of tasks that succeeded
`percent_finished`	float	Fraction of tasks that completed (success or fail)
`percent_error`	float	Fraction of tasks that errored
`total_agent_cost`	float	Total cost across all tasks (USD)
`planned_sessions` / `total_sessions` / `successful_sessions`	int	Session counts

Downloads last month: 164

Space using open-agent-leaderboard/results 1

Collection including open-agent-leaderboard/results

Open Agent Leaderboard

Collection

An open benchmark for comparing full agent systems across diverse real-world tasks. Reports both quality and cost. • 4 items • Updated Mar 30 • 1

Paper for open-agent-leaderboard/results

General Agent Evaluation

Paper • 2602.22953 • Published Feb 26 • 12