[🤗 HF Leaderboard] [📄 Paper]

🎉 Updates

• 2025/5/23: The paper "Unifying Language Agent Algorithms with Graph-based Orchestration Engine for Reproducible Agent Research" was accepted to ACL 2025 Systems Demonstration Track.🎉
• 2025/2/11: Add deepseek-r1:1.5b, a new dataset MATH-500, and a new algorithm ToT into the leaderboard.
• 2025/1/23: Add gpt-4o, Qwen2.5-72B-Instruct, Qwen2.5-7B-Instruct, Qwen2-1.5B-Instruct, Qwen2-0.5B-Instruct, Llama-3.3-70B-Instruct, Llama-3.1-8B-Instruct, Internllm2_5-7B into the leaderboard.
• 2025/1/07: The Open Agent Leaderboard is released.

📖 Introduction

This project aims to provide a fair comparison of various agents by evaluating their performance on different datasets and LLMs. Built on top of the OmAgent framework, it allows for simple, quick, and accurate assessments of agents.

Supported benchmark datasets:

• gsm8k
• AQuA
• MATH-500

Supported algorithms:

• IO: Input-Output Direct Prompting (Baseline)
• CoT: Chain-of-thought prompting elicits reasoning in large language models, Large Language Models are Zero-Shot Reasoners
• SC-CoT: Self-Consistency Improves Chain of Thought Reasoning in Language Models
• PoT: Program of thoughts prompting: Disentangling computation from reasoning for numerical reasoning tasks
• ReAct: ReAct: Synergizing Reasoning and Acting in Language Models
• ToT: Tree of Thoughts: Deliberate Problem Solving with Large Language Models

Supported LLMs:

• gpt-3.5-turbo
• gpt-4o
• Doubao-lite-32k
• Qwen2.5-72B-Instruct
• Qwen2.5-7B-Instruct
• Qwen2-1.5B-Instruct
• Qwen2-0.5B-Instruct
• Llama-3.3-70B-Instruct
• Llama-3.1-8B-Instruct
• Internllm2_5-7B
• deepseek-r1:1.5b

🌟 Graph-based Workflow Orchestration Engine

At the core of AGORA is a graph-based orchestration engine designed for modularity and scalability. As shown in the figure below, the system uses a Directed Acyclic Graph (DAG) where each node represents a task. Tasks are either simple tasks—developer-defined custom logic—or logical tasks—built-in control flows such as branching and looping.

Built on the Conductor library, this engine provides visual representations of workflows, making agent behavior intuitive to trace and debug. It also supports asynchronous, distributed execution, which is ideal for managing long-running, complex agent workflows.

🏅 Leaderboards

Math tasks

Rank	Algorithm	LLM	Eval Date	Avg Score	gsm8k-Score	gsm8k-Cost($)	AQuA-Score	AQuA-Cost($)	MATH-500-Score	MATH-500-Cost($)
1	SC-CoT	Qwen2.5-72B-Instruct	2025/1/22	86.67	94.77	4.045	85.43	0.4186	79.8	1.8504
2	CoT	Qwen2.5-72B-Instruct	2025/1/22	86.43	92.87	0.7195	86.22	0.0808	80.2	0.349
3	SC-CoT	gpt-4o	2025/1/22	85.07	94.77	18.2044	85.83	5.2456	74.6	12.3611
4	SC-CoT	Llama-3.3-70B-Instruct	2025/1/22	84.09	95.22	3.7895	84.65	0.4438	72.4	1.7845
5	CoT	Llama-3.3-70B-Instruct	2025/1/22	82.86	93.93	0.687	83.46	0.0927	71.2	0.3463
6	CoT	gpt-4o	2025/1/22	81.59	94.09	4.5367	82.68	1.0417	68	3.0569
7	IO	Llama-3.3-70B-Instruct	2025/1/22	81.45	92.27	0.4709	82.68	0.0798	69.4	0.2386
8	SC-CoT	Qwen2.5-7B-Instruct	2025/1/22	80.57	90.98	0	79.53	0	71.2	0
9	IO	Qwen2.5-72B-Instruct	2025/1/22	80.34	86.58	0.4899	84.25	0.0742	70.2	0.2506
10	CoT	Qwen2.5-7B-Instruct	2025/1/22	78.73	85.67	0	80.71	0	69.8	0
11	SC-CoT	Doubao-lite-32k	2025/1/7	77.92	91.58	0.1118	76.37	0.0279	65.8	0.0734
12	ReAct-Pro*	Llama-3.3-70B-Instruct	2025/1/22	77.12	87.64	10.1124	79.13	0.768	64.6	3.1806
13	CoT	Doubao-lite-32k	2025/1/7	77	89.31	0.0558	82.68	0.0066	59	0.0255
14	ReAct-Pro*	Qwen2.5-72B-Instruct	2025/1/22	74.43	87.26	10.5479	73.23	0.3177	62.8	3.4541
15	PoT	Qwen2.5-72B-Instruct	2025/1/22	71.58	92.34	0.7054	75.2	0.1645	47.2	0.233
16	PoT	gpt-4o	2025/1/22	71.5	93.1	4.2166	75.2	1.6087	46.2	1.5994
17	ReAct-Pro*	Doubao-lite-32k	2025/1/7	70.12	85.6	0.2512	77.56	0.0445	47.2	0.186
18	ReAct-Pro*	Qwen2.5-7B-Instruct	2025/1/22	68.69	82.87	0	74.41	0	48.8	0
19	IO	gpt-4o	2025/1/22	68.6	88.4	3.3463	75.59	1.1453	41.8	2.7907
20	IO	Qwen2.5-7B-Instruct	2025/1/22	65.13	57.24	0	78.74	0	59.4	0
21	PoT	Llama-3.3-70B-Instruct	2025/1/22	65.07	73.09	0.9736	79.53	0.1746	42.6	0.2839
22	CoT	deepseek-r1:1.5b	2025/1/23	63.9	70.66	0	71.65	0	49.4	0
23	IO	Doubao-lite-32k	2025/1/7	62.85	72.02	0.0354	79.13	0.0058	37.4	0.0187
24	PoT	Doubao-lite-32k	2025/1/7	61.29	79.61	0.0576	71.65	0.0147	32.6	0.0144
25	ToT	Qwen2.5-72B-Instruct	2025/1/22	60.26	88.88	23.5911	81.1	3.7389	10.8	9.0421
26	CoT	gpt-3.5-turbo	2025/1/7	59.84	78.7	0.6788	61.02	0.0957	39.8	0.3189
27	CoT	Internllm2_5-7B	2025/1/22	59.02	77.71	0	52.76	0	46.6	0
28	IO	deepseek-r1:1.5b	2025/1/22	58.95	64.14	0	68.9	0	43.8	0
29	ToT	Llama-3.3-70B-Instruct	2025/1/22	58.79	91.89	20.8753	83.07	2.9404	1.4	8.2699
30	ToT	gpt-4o	2025/1/22	58.61	91.13	86.8581	81.5	8.5295	3.2	40.8094
31	ReAct-Pro*	gpt-4o	2025/1/22	58.26	63.31	39.0751	57.48	2.304	54	17.7735
32	SC-CoT	deepseek-r1:1.5b	2025/2/10	57.91	69.07	0	57.87	0	46.8	0
33	SC-CoT	gpt-3.5-turbo	2025/1/7	56.25	69.29	2.5203	58.66	0.3277	40.8	1.2308
34	PoT	Qwen2.5-7B-Instruct	2025/1/22	55.51	58.83	0	68.11	0	39.6	0
35	PoT	gpt-3.5-turbo	2025/1/7	55.04	76.88	0.6902	59.45	0.1748	28.8	0.168
36	ReAct-Pro*	gpt-3.5-turbo	2025/1/7	54.43	74.91	3.4633	64.57	0.4928	23.8	2.0406
37	CoT	Llama-3.1-8B-Instruct	2025/1/22	53.96	75.44	0	60.63	0	25.8	0
38	ReAct-Pro*	Llama-3.1-8B-Instruct	2025/1/22	50.7	67.78	0	55.51	0	28.8	0
39	IO	Llama-3.1-8B-Instruct	2025/1/22	48.98	57.16	0	51.18	0	38.6	0
40	ToT	gpt-3.5-turbo	2025/1/7	44.94	67.93	9.1707	57.09	1.1513	9.8	5.2914
41	SC-CoT	Llama-3.1-8B-Instruct	2025/1/22	44.54	54.36	0	59.45	0	19.8	0
42	ToT	Qwen2.5-7B-Instruct	2025/1/22	42.52	72.21	0	53.94	0	1.4	0
43	ToT	Llama-3.1-8B-Instruct	2025/1/22	41.97	65.05	0	59.06	0	1.8	0
44	ReAct-Pro*	deepseek-r1:1.5b	2025/2/10	38.22	35.94	0	54.33	0	24.4	0
45	CoT	Qwen2-1.5B-Instruct	2025/1/22	37.08	55.5	0	40.55	0	15.2	0
46	PoT	Llama-3.1-8B-Instruct	2025/1/22	33.56	38.67	0	36.61	0	25.4	0
47	IO	gpt-3.5-turbo	2025/1/7	31.34	37.83	0.3328	38.98	0.038	17.2	0.2436
48	SC-CoT	Internllm2_5-7B	2025/1/22	30.81	44.66	0	38.58	0	9.2	0
49	PoT	Internllm2_5-7B	2025/1/22	29.94	38.21	0	36.61	0	15	0
50	ReAct-Pro*	Internllm2_5-7B	2025/1/22	29.75	33.51	0	40.94	0	14.8	0
51	ToT	Doubao-lite-32k	2025/1/7	28.1	37.83	0.8739	45.28	0.0881	1.2	0.2371
52	IO	Internllm2_5-7B	2025/1/22	27.35	11.6	0	47.64	0	22.8	0
53	CoT	Qwen2-0.5B-Instruct	2025/1/22	25.07	35.94	0	33.07	0	6.2	0
54	PoT	deepseek-r1:1.5b	2025/2/10	22.54	11.9	0	54.72	0	1	0
55	ReAct-Pro*	Qwen2-1.5B-Instruct	2025/1/22	19.55	24.87	0	25.59	0	8.2	0
56	ToT	Internllm2_5-7B	2025/1/22	18.96	20.85	0	35.83	0	0.2	0
57	IO	Qwen2-1.5B-Instruct	2025/1/22	17.6	16.68	0	29.13	0	7	0
58	ToT	Qwen2-1.5B-Instruct	2025/1/22	17.31	19.64	0	31.5	0	0.8	0
59	PoT	Qwen2-1.5B-Instruct	2025/1/22	16.67	18.5	0	30.71	0	0.8	0
60	ToT	deepseek-r1:1.5b	2025/2/10	16.11	23.12	0	24.8	0	0.4	0
61	IO	Qwen2-0.5B-Instruct	2025/1/22	14.83	14.71	0	27.17	0	2.6	0
62	ReAct-Pro*	Qwen2-0.5B-Instruct	2025/1/22	10.76	7.66	0	24.02	0	0.6	0
63	ToT	Qwen2-0.5B-Instruct	2025/1/22	9.97	0	0	29.92	0	0	0
64	PoT	Qwen2-0.5B-Instruct	2025/1/22	8.98	9.63	0	17.32	0	0	0
65	SC-CoT	Qwen2-0.5B-Instruct	2025/1/22	7.9	4.17	0	17.32	0	2.2	0
66	SC-CoT	Qwen2-1.5B-Instruct	2025/1/22	6.94	8.19	0	10.63	0	2	0

Evaluation details can be found in the Evaluation Details section and huggingface leaderboard.

• IO (Input-Output) is the baseline method that directly prompts the model with the question and expects an answer without any intermediate reasoning steps. It represents the most basic way of using language models and serves as a reference point for evaluating the effectiveness of other algorithms.

• ReAct-Pro*: We modified ReAct to ReAct-Pro, following the Reflexion repository. Comparasion with the original ReAct repo can be found in the Compare to ReAct section.

🛠️ How to Install

1. Clone the repository:

   git clone https://github.com/om-ai-lab/open-agent-leaderboard.git
   cd open-agent-leaderboard
   

2. Install dependencies:

   pip install -r requirements.txt
   

🏗️ How to Evaluate Agents

Step 1. Implement your agent in the omagent repository

Navigate to the agent repository:

git clone https://github.com/om-ai-lab/OmAgent.git
cd OmAgent

Set up the environment:

pip install -e omagent-core

Implement your agent in the omagent repository, check the examples/cot folder.

Step 2. Inference in OmAgent Repository

Run the inference script (cot as an example):

cd examples/cot
python eval_demo.py --model_id your_model_id --dataset_name your_dataset_name --dataset_path your_dataset_path --output_path your_output_path --output_name your_output_name --cot_method your_cot_method

Output Format

The output results are saved in JSON format and include the following fields:

• id: The unique identifier of the sample.
• question: The input question provided to the model.
• last_output: The raw output generated by the model.
• output_postprocess (optional): The processed output after cleansing.
• ground_truth (optional): The correct answer for the sample.
• prompt_tokens: The number of tokens in the input prompt.
• completion_tokens: The number of tokens in the model's output.

Example of an output JSON file:

{
  "dataset": "gsm8k",
  "model_id": "gpt-3.5-turbo",
  "alg": "COT",
  "model_result": [
    {
      "id": 1,
      "question": "Q: There are 15 trees in the grove. Grove workers will plant trees in the grove today.....",
      "last_output": "Janet's ducks lay 16 eggs per day. She eats 3 for breakfast and uses 4 to bake muffins,...",
      "output_postprocess": "Paris",
      "ground_truth": "Paris",
      "prompt_tokens": 10,
      "completion_tokens": 5
    }
  ]
}

Step 3. Evaluate inference results

Run the main script to perform evaluations:

python main.py --dataset <dataset_name> --model <model_name> --method <method_name> --output_dir <output_directory>

Parameters

• --random_seed: Random seed, default is 1.
• --dataset: Dataset to use, options are aqua, gsm8k, math500.
• --minibatch_size: Minibatch size, default is 1.
• --max_num_worker: Maximum number of workers for the data loader, default is 4.
• --model: Model used for decoding, options are gpt-4o-mini, gpt-4o, gpt-3.5-turbo.
• --method: Method, options are zero_shot, zero_shot_cot, few_shot, few_shot_cot.
• --cot_trigger_no: Trigger sentence number for chain of thought, default is 1.
• --max_length: Maximum length of model output, default is 2048.
• --max_length_direct: Maximum length of direct model answer, default is 32.
• --limit_dataset_size: Whether to limit the test dataset size, default is 0 (no limit).
• --output_dir: Output directory, default is ./outputs/.
• --output_path: Output path, default is empty.
• --agent: Agent used for the experiment, options are cot, pot, sc_cot, react.
• --system_prompt: System prompt, default is empty.
• --openai_api_key: OpenAI API key, default is empty.
• --openai_url: OpenAI API URL, default is https://api.openai.com/v1.

Example

python main.py --output_path example/gsm8k_results_cot.json --dataset gsm8k --method few_shot_cot

Evaluation details

Algorithm	Dataset	Eval Date	LLM	Score	Pass rate	X-shot	Parameters	Samples	Total input tokens	Average input tokens	Total output tokens	Average output tokens	All tokens	Cost($)
IO	gsm8k	2025/1/7	gpt-3.5-turbo	37.83	99.92	8		1,319	546,990	415	39,563	30	586,553	0.3328
IO	gsm8k	2025/1/7	Doubao-lite-32k	72.02	99.92	8		1,319	617,377	468	123,106	93	740,483	0.0354
IO	gsm8k	2025/1/22	gpt-4o	88.4	100	8		1,319	542,416	411	199,030	151	741,446	3.3463
IO	gsm8k	2025/1/22	Qwen2.5-72B-Instruct	86.58	100	8		1,319	555,340	421	313,720	238	869,060	0.4899
IO	gsm8k	2025/1/22	Llama-3.3-70B-Instruct	92.27	100	8		1,319	583,916	443	251,359	191	835,275	0.4709
IO	gsm8k	2025/1/22	Qwen2.5-7B-Instruct	57.24	100	8		1,319	596,229	452	291,684	221	887,913	0
IO	gsm8k	2025/1/22	Llama-3.1-8B-Instruct	57.16	99.55	8		1,319	550,941	418	1,194,488	906	1,745,429	0
IO	gsm8k	2025/1/22	Internllm2_5-7B	11.6	97.95	8		1,319	679,302	515	434,426	329	1,113,728	0
IO	gsm8k	2025/1/22	Qwen2-1.5B-Instruct	16.68	100	8		1,319	568,530	431	168,466	128	736,996	0
IO	gsm8k	2025/1/22	Qwen2-0.5B-Instruct	14.71	100	8		1,319	568,116	431	266,781	202	834,897	0
IO	gsm8k	2025/1/22	deepseek-r1:1.5b	64.14	99.62	8		1,319	561,935	426	921,116	698	1,483,051	0
ReAct-Pro*	gsm8k	2025/1/7	gpt-3.5-turbo	74.91	99.39	8	max_steps=10	1,319	6,506,164	4,933	140,122	106	6,646,286	3.4633
ReAct-Pro*	gsm8k	2025/1/7	Doubao-lite-32k	85.6	99.62	8	max_steps=10	1,319	5,862,016	4,444	136,623	104	5,998,639	0.2512
ReAct-Pro*	gsm8k	2025/1/22	gpt-4o	63.31	99.55	8	max_steps=10	1,319	14,411,173	10,926	304,714	231	14,715,887	39.0751
ReAct-Pro*	gsm8k	2025/1/22	Qwen2.5-72B-Instruct	87.26	100	8	max_steps=10	1,319	18,160,983	13,769	549,454	417	18,710,437	10.5479
ReAct-Pro*	gsm8k	2025/1/22	Llama-3.3-70B-Instruct	87.64	99.92	8	max_steps=10	1,319	17,038,928	12,918	898,936	682	17,937,864	10.1124
ReAct-Pro*	gsm8k	2025/1/22	Qwen2.5-7B-Instruct	82.87	100	8	max_steps=10	1,319	14,355,752	10,884	495,162	375	14,850,914	0
ReAct-Pro*	gsm8k	2025/1/22	Llama-3.1-8B-Instruct	67.78	98.56	8	max_steps=10	1,319	21,044,978	15,955	1,790,789	1,358	22,835,767	0
ReAct-Pro*	gsm8k	2025/1/22	Internllm2_5-7B	33.51	97.95	8	max_steps=10	1,319	30,120,070	22,836	5,549,919	4,208	35,669,989	0
ReAct-Pro*	gsm8k	2025/1/22	Qwen2-1.5B-Instruct	24.87	80.21	8	max_steps=10	1,319	9,133,603	6,925	694,398	526	9,828,001	0
ReAct-Pro*	gsm8k	2025/1/22	Qwen2-0.5B-Instruct	7.66	95.22	8	max_steps=10	1,319	52,431,343	39,751	2,961,268	2,245	55,392,611	0
ReAct-Pro*	gsm8k	2025/2/10	deepseek-r1:1.5b	35.94	99.62	8	max_steps=10	1,319	19,299,381	14,632	4,919,696	3,730	24,219,077	0
PoT	gsm8k	2025/1/7	gpt-3.5-turbo	76.88	99.24	8		1,319	1,090,418	827	96,662	73	1,187,080	0.6902
PoT	gsm8k	2025/1/7	Doubao-lite-32k	79.61	92.57	8		1,319	1,170,038	887	118,017	89	1,288,055	0.0576
PoT	gsm8k	2025/1/22	gpt-4o	93.1	99.77	8		1,319	1,101,672	835	146,240	111	1,247,912	4.2166
PoT	gsm8k	2025/1/22	Qwen2.5-72B-Instruct	92.34	99.39	8		1,319	1,106,682	839	144,528	110	1,251,210	0.7054
PoT	gsm8k	2025/1/22	Llama-3.3-70B-Instruct	73.09	79.61	8		1,319	1,126,025	854	601,019	456	1,727,044	0.9736
PoT	gsm8k	2025/1/22	Qwen2.5-7B-Instruct	58.83	70.51	8		1,319	1,145,390	868	217,432	165	1,362,822	0
PoT	gsm8k	2025/1/22	Llama-3.1-8B-Instruct	38.67	55.42	8		1,319	1,147,538	870	243,573	185	1,391,111	0
PoT	gsm8k	2025/1/22	Internllm2_5-7B	38.21	48.9	8		1,319	1,136,843	862	188,106	143	1,324,949	0
PoT	gsm8k	2025/1/22	Qwen2-1.5B-Instruct	18.5	31.01	8		1,319	1,151,528	873	175,994	133	1,327,522	0
PoT	gsm8k	2025/1/22	Qwen2-0.5B-Instruct	9.63	16.91	8		1,319	1,151,528	873	237,607	180	1,389,135	0
PoT	gsm8k	2025/2/10	deepseek-r1:1.5b	11.9	17.44	8		1,319	1,138,872	863	815,637	618	1,954,509	0
CoT	gsm8k	2025/1/7	gpt-3.5-turbo	78.7	100	8		1,319	953,242	723	134,799	102	1,088,041	0.6788
CoT	gsm8k	2025/1/7	Doubao-lite-32k	89.31	100	8		1,319	1,042,095	790	159,725	121	1,201,820	0.0558
CoT	gsm8k	2025/1/22	gpt-4o	94.09	100	8		1,319	948,668	719	216,498	164	1,165,166	4.5367
CoT	gsm8k	2025/1/22	Qwen2.5-72B-Instruct	92.87	100	8		1,319	1,005,119	762	271,133	206	1,276,252	0.7195
CoT	gsm8k	2025/1/22	Llama-3.3-70B-Instruct	93.93	100	8		1,319	990,168	751	228,497	173	1,218,665	0.687
CoT	gsm8k	2025/1/22	Qwen2.5-7B-Instruct	85.67	100	8		1,319	1,046,008	793	244,797	186	1,290,805	0
CoT	gsm8k	2025/1/22	Llama-3.1-8B-Instruct	75.44	99.92	8		1,319	990,168	751	258,161	196	1,248,329	0
CoT	gsm8k	2025/1/22	Internllm2_5-7B	77.71	99.7	8		1,319	968,163	734	234,000	177	1,202,163	0
CoT	gsm8k	2025/1/22	Qwen2-1.5B-Instruct	55.5	100	8		1,319	1,032,818	783	185,707	141	1,218,525	0
CoT	gsm8k	2025/1/22	Qwen2-0.5B-Instruct	35.94	99.92	8		1,319	1,032,818	783	190,641	145	1,223,459	0
CoT	gsm8k	2025/1/23	deepseek-r1:1.5b	70.66	99.77	8		1,319	1,011,714	767	1,078,911	818	2,090,625	0
SC-CoT	gsm8k	2025/1/7	gpt-3.5-turbo	69.29	98.79	8	temperature=1, path_num=5	1,319	895,571	679	1,381,678	1,048	2,277,249	2.5203
SC-CoT	gsm8k	2025/1/7	Doubao-lite-32k	91.58	99.92	8	temperature=1, path_num=5	1,319	942,182	714	893,709	678	1,835,891	0.1118
SC-CoT	gsm8k	2025/1/22	gpt-4o	94.77	100	8	temperature=1, path_num=5	1,319	894,889	678	1,596,716	1,211	2,491,605	18.2044
SC-CoT	gsm8k	2025/1/22	Qwen2.5-72B-Instruct	94.77	100	8	temperature=1, path_num=5	1,319	5,370,360	4,072	1,804,898	1,368	7,175,258	4.045
SC-CoT	gsm8k	2025/1/22	Llama-3.3-70B-Instruct	95.22	100	8	temperature=1, path_num=5	1,319	5,295,585	4,015	1,426,429	1,081	6,722,014	3.7895
SC-CoT	gsm8k	2025/1/22	Qwen2.5-7B-Instruct	90.98	100	8	temperature=1, path_num=5	1,319	5,580,524	4,231	1,679,419	1,273	7,259,943	0
SC-CoT	gsm8k	2025/1/22	Llama-3.1-8B-Instruct	54.36	99.85	8	temperature=1, path_num=5	1,319	5,136,762	3,894	5,819,672	4,412	10,956,434	0
SC-CoT	gsm8k	2025/1/22	Internllm2_5-7B	44.66	91.81	8	temperature=1, path_num=5	1,319	5,847,761	4,433	2,314,738	1,755	8,162,499	0
SC-CoT	gsm8k	2025/1/22	Qwen2-1.5B-Instruct	8.19	68.76	8	temperature=1, path_num=5	1,319	5,439,568	4,124	1,946,885	1,476	7,386,453	0
SC-CoT	gsm8k	2025/1/22	Qwen2-0.5B-Instruct	4.17	94.47	8	temperature=1, path_num=5	1,319	5,441,962	4,126	2,036,805	1,544	7,478,767	0
SC-CoT	gsm8k	2025/2/10	deepseek-r1:1.5b	69.07	98.79	8	temperature=1, path_num=5	1,319	5,407,357	4,100	4,622,327	3,504	10,029,684	0
ToT	gsm8k	2025/1/7	gpt-3.5-turbo	67.93	99.7	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	15,920,037	12,070	807,138	612	16,727,175	9.1707
ToT	gsm8k	2025/1/7	Doubao-lite-32k	37.83	87.34	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	19,208,597	14,563	1,065,752	808	20,274,349	0.8739
ToT	gsm8k	2025/1/22	gpt-4o	91.13	100	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	29,445,237	22,324	1,324,498	1,004	30,769,735	86.8581
ToT	gsm8k	2025/1/22	Qwen2.5-72B-Instruct	88.88	100	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	40,435,361	30,656	1,411,787	1,070	41,847,148	23.5911
ToT	gsm8k	2025/1/22	Llama-3.3-70B-Instruct	91.89	100	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	35,096,810	26,609	1,932,877	1,465	37,029,687	20.8753
ToT	gsm8k	2025/1/22	Qwen2.5-7B-Instruct	72.21	99.01	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	20,196,528	15,312	11,460,791	8,689	31,657,319	0
ToT	gsm8k	2025/1/22	Llama-3.1-8B-Instruct	65.05	91.96	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	15,554,967	11,793	877,135	665	16,432,102	0
ToT	gsm8k	2025/1/22	Internllm2_5-7B	20.85	70.13	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	11,768,118	8,922	1,410,011	1,069	13,178,129	0
ToT	gsm8k	2025/1/22	Qwen2-1.5B-Instruct	19.64	77.26	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	12,124,248	9,192	634,439	481	12,758,687	0
ToT	gsm8k	2025/1/22	Qwen2-0.5B-Instruct	-	-	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	-	-	-	-	-	-
ToT	gsm8k	2025/2/10	deepseek-r1:1.5b	23.12	72.48	8	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	1,319	2,738,244	2,076	683,242	518	3,421,486	0
IO	AQuA	2025/1/7	gpt-3.5-turbo	38.98	100	0		254	25,701	101	16,770	66	42,471	0.038
IO	AQuA	2025/1/7	Doubao-lite-32k	79.13	100	0		254	33,058	130	54,684	215	87,742	0.0058
IO	AQuA	2025/1/22	gpt-4o	75.59	97.24	0		254	25,631	101	108,121	426	133,752	1.1453
IO	AQuA	2025/1/22	Qwen2.5-72B-Instruct	84.25	99.61	0		254	25,397	100	106,207	418	131,604	0.0742
IO	AQuA	2025/1/22	Llama-3.3-70B-Instruct	82.68	99.21	0		254	32,809	129	108,758	428	141,567	0.0798
IO	AQuA	2025/1/22	Qwen2.5-7B-Instruct	78.74	98.43	0		254	33,271	131	104,500	411	137,771	0
IO	AQuA	2025/1/22	Llama-3.1-8B-Instruct	51.18	98.82	0		254	26,459	104	106,647	420	133,106	0
IO	AQuA	2025/1/22	Internllm2_5-7B	47.64	90.94	0		254	50,232	198	134,809	531	185,041	0
IO	AQuA	2025/1/22	Qwen2-1.5B-Instruct	29.13	97.64	0		254	27,937	110	43,110	170	71,047	0
IO	AQuA	2025/1/22	Qwen2-0.5B-Instruct	27.17	98.82	0		254	27,937	110	82,478	325	110,415	0
IO	AQuA	2025/1/22	deepseek-r1:1.5b	68.9	94.88	0		254	26,667	105	325,100	1,280	351,767	0
CoT	AQuA	2025/1/7	gpt-3.5-turbo	61.02	93.7	0		254	25,447	100	55,346	218	80,793	0.0957
CoT	AQuA	2025/1/7	Doubao-lite-32k	82.68	97.24	0		254	27,978	110	66,599	262	94,577	0.0066
CoT	AQuA	2025/1/22	gpt-4o	82.68	98.03	0		254	25,123	99	97,894	385	123,017	1.0417
CoT	AQuA	2025/1/22	Qwen2.5-72B-Instruct	86.22	99.21	0		254	25,143	99	118,146	465	143,289	0.0808
CoT	AQuA	2025/1/22	Llama-3.3-70B-Instruct	83.46	98.43	0		254	32,555	128	131,834	519	164,389	0.0927
CoT	AQuA	2025/1/22	Qwen2.5-7B-Instruct	80.71	99.61	0		254	33,017	130	116,719	460	149,736	0
CoT	AQuA	2025/1/22	Llama-3.1-8B-Instruct	60.63	100	0		254	32,555	128	111,880	440	144,435	0
CoT	AQuA	2025/1/22	Internllm2_5-7B	52.76	89.37	0		254	26,610	105	100,910	397	127,520	0
CoT	AQuA	2025/1/22	Qwen2-1.5B-Instruct	40.55	98.82	0		254	30,477	120	79,563	313	110,040	0
CoT	AQuA	2025/1/22	Qwen2-0.5B-Instruct	33.07	98.82	0		254	30,477	120	86,862	342	117,339	0
CoT	AQuA	2025/1/23	deepseek-r1:1.5b	71.65	96.85	0		254	26,413	104	306,659	1,207	333,072	0
PoT	AQuA	2025/1/7	gpt-3.5-turbo	59.45	100	0		254	225,162	886	41,492	163	266,654	0.1748
PoT	AQuA	2025/1/7	Doubao-lite-32k	71.65	96.85	0		254	259,863	1,023	49,573	195	309,436	0.0147
PoT	AQuA	2025/1/22	gpt-4o	75.2	100	0		254	222,717	877	105,191	414	327,908	1.6087
PoT	AQuA	2025/1/22	Qwen2.5-72B-Instruct	75.2	100	0		254	249,215	981	42,549	168	291,764	0.1645
PoT	AQuA	2025/1/22	Llama-3.3-70B-Instruct	79.53	99.21	0		254	240,735	948	69,064	272	309,799	0.1746
PoT	AQuA	2025/1/22	Qwen2.5-7B-Instruct	68.11	100	0		254	264,517	1,041	49,211	194	313,728	0
PoT	AQuA	2025/1/22	Llama-3.1-8B-Instruct	36.61	96.85	0		254	240,613	947	50,301	198	290,914	0
PoT	AQuA	2025/1/22	Internllm2_5-7B	36.61	98.82	0		254	233,505	919	68,457	270	301,962	0
PoT	AQuA	2025/1/22	Qwen2-1.5B-Instruct	30.71	96.46	0		254	246,560	971	51,915	204	298,475	0
PoT	AQuA	2025/1/22	Qwen2-0.5B-Instruct	17.32	92.13	0		254	258,867	1,019	63,414	250	322,281	0
PoT	AQuA	2025/2/10	deepseek-r1:1.5b	54.72	97.24	0		254	250,690	987	765,957	3,016	1,016,647	0
SC-CoT	AQuA	2025/1/22	gpt-3.5-turbo	58.66	92.52	0	temperature=1, path_num=5	254	27,906	110	209,160	823	237,066	0.3277
SC-CoT	AQuA	2025/1/22	Doubao-lite-32k	76.37	91.73	0	temperature=1, path_num=5	254	31,703	125	325,136	1,280	356,839	0.0279
SC-CoT	AQuA	2025/1/22	gpt-4o	85.83	99.21	0	temperature=1, path_num=5	254	27,829	110	517,602	2,038	545,431	5.2456
SC-CoT	AQuA	2025/1/22	Qwen2.5-72B-Instruct	85.43	96.85	0	temperature=1, path_num=5	254	137,990	543	604,562	2,380	742,552	0.4186
SC-CoT	AQuA	2025/1/22	Llama-3.3-70B-Instruct	84.65	99.61	0	temperature=1, path_num=5	254	175,050	689	612,262	2,410	787,312	0.4438
SC-CoT	AQuA	2025/1/22	Qwen2.5-7B-Instruct	79.53	100	0	temperature=1, path_num=5	254	177,972	701	567,438	2,234	745,410	0
SC-CoT	AQuA	2025/1/22	Llama-3.1-8B-Instruct	59.45	95.67	0	temperature=1, path_num=5	254	145,108	571	544,969	2,146	690,077	0
SC-CoT	AQuA	2025/1/22	Internllm2_5-7B	38.58	97.24	0	temperature=1, path_num=5	254	264,557	1,042	615,114	2,422	879,671	0
SC-CoT	AQuA	2025/1/22	Qwen2-1.5B-Instruct	10.63	51.57	0	temperature=1, path_num=5	254	151,410	596	550,570	2,168	701,980	0
SC-CoT	AQuA	2025/1/22	Qwen2-0.5B-Instruct	17.32	82.28	0	temperature=1, path_num=5	254	150,787	594	603,126	2,375	753,913	0
SC-CoT	AQuA	2025/2/10	deepseek-r1:1.5b	57.87	74.02	0	temperature=1, path_num=5	254	144,710	570	1,987,401	7,824	2,132,111	0
ReAct-Pro*	AQuA	2025/1/7	gpt-3.5-turbo	64.57	98.03	0	max_steps=10	254	862,614	3,396	40,973	161	903,587	0.4928
ReAct-Pro*	AQuA	2025/1/7	Doubao-lite-32k	77.56	96.06	0	max_steps=10	254	977,890	3,850	54,951	216	1,032,841	0.0445
ReAct-Pro*	AQuA	2025/1/22	gpt-4o	57.48	97.24	0	max_steps=10	254	615,589	2,424	76,507	301	692,096	2.304
ReAct-Pro*	AQuA	2025/1/22	Qwen2.5-72B-Instruct	73.23	100	0	max_steps=10	254	441,765	1,739	121,838	480	563,603	0.3177
ReAct-Pro*	AQuA	2025/1/22	Llama-3.3-70B-Instruct	79.13	99.61	0	max_steps=10	254	1,119,143	4,406	243,236	958	1,362,379	0.768
ReAct-Pro*	AQuA	2025/1/22	Qwen2.5-7B-Instruct	74.41	99.21	0	max_steps=10	254	564,165	2,221	131,679	518	695,844	0
ReAct-Pro*	AQuA	2025/1/22	Llama-3.1-8B-Instruct	55.51	96.85	0	max_steps=10	254	3,764,723	14,822	576,098	2,268	4,340,821	0
ReAct-Pro*	AQuA	2025/1/22	Internllm2_5-7B	40.94	96.85	0	max_steps=10	254	3,592,039	14,142	836,762	3,294	4,428,801	0
ReAct-Pro*	AQuA	2025/1/22	Qwen2-1.5B-Instruct	25.59	96.06	0	max_steps=10	254	4,555,858	17,936	516,146	2,032	5,072,004	0
ReAct-Pro*	AQuA	2025/1/22	Qwen2-0.5B-Instruct	24.02	96.85	0	max_steps=10	254	6,344,167	24,977	825,920	3,252	7,170,087	0
ReAct-Pro*	AQuA	2025/2/10	deepseek-r1:1.5b	54.33	96.46	0	max_steps=10	254	10,578,715	41,648	3,866,326	15,222	14,445,041	0
ToT	AQuA	2025/1/7	gpt-3.5-turbo	57.09	99.61	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	1,850,767	7,286	150,629	593	2,001,396	1.1513
ToT	AQuA	2025/1/7	Doubao-lite-32k	45.28	74.02	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	1,850,249	7,284	150,301	592	2,000,550	0.0881
ToT	AQuA	2025/1/22	gpt-4o	81.5	99.21	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	2,347,538	9,242	266,069	1,048	2,613,607	8.5295
ToT	AQuA	2025/1/22	Qwen2.5-72B-Instruct	81.1	99.21	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	6,371,642	25,085	260,613	1,026	6,632,255	3.7389
ToT	AQuA	2025/1/22	Llama-3.3-70B-Instruct	83.07	100	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	4,735,188	18,642	480,660	1,892	5,215,848	2.9404
ToT	AQuA	2025/1/22	Qwen2.5-7B-Instruct	53.94	100	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	8,224,468	32,380	378,214	1,489	8,602,682	0
ToT	AQuA	2025/1/22	Llama-3.1-8B-Instruct	59.06	100	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	4,896,222	19,276	843,462	3,321	5,739,684	0
ToT	AQuA	2025/1/22	Internllm2_5-7B	35.83	99.61	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	4,263,136	16,784	471,424	1,856	4,734,560	0
ToT	AQuA	2025/1/22	Qwen2-1.5B-Instruct	31.5	98.82	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	6,058,022	23,850	192,680	759	6,250,702	0
ToT	AQuA	2025/1/22	Qwen2-0.5B-Instruct	29.92	100	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	8,100,085	31,890	600,196	2,363	8,700,281	0
ToT	AQuA	2025/2/10	deepseek-r1:1.5b	24.8	55.51	0	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	254	605,028	2,382	189,484	746	794,512	0
IO	MATH-500	2025/1/24	gpt-3.5-turbo	17.2	100	4		500	154,881	310	110,744	221	265,625	0.2436
IO	MATH-500	2025/1/24	Doubao-lite-32k	37.4	100	4		500	166,870	334	144,860	290	311,730	0.0187
IO	MATH-500	2025/1/22	gpt-4o	41.8	100	4		500	153,832	308	240,615	481	394,447	2.7907
IO	MATH-500	2025/1/24	Qwen2.5-72B-Instruct	70.2	100	4		500	169,549	339	275,042	550	444,591	0.2506
IO	MATH-500	2025/1/24	Llama-3.3-70B-Instruct	69.4	100	4		500	155,879	312	267,337	535	423,216	0.2386
IO	MATH-500	2025/1/24	Qwen2.5-7B-Instruct	59.4	100	4		500	169,549	339	241,813	484	411,362	0
IO	MATH-500	2025/1/24	Llama-3.1-8B-Instruct	38.6	100	4		500	155,563	311	348,371	697	503,934	0
IO	MATH-500	2025/1/24	Internllm2_5-7B	22.8	100	4		500	201,883	404	266,005	532	467,888	0
IO	MATH-500	2025/1/24	Qwen2-1.5B-Instruct	7	100	4		500	158,777	318	255,101	510	413,878	0
IO	MATH-500	2025/1/24	Qwen2-0.5B-Instruct	2.6	100	4		500	159,049	318	270,281	541	429,330	0
IO	MATH-500	2025/1/24	deepseek-r1:1.5b	43.8	100	4		500	157,049	314	865,499	1,731	1,022,548	0
CoT	MATH-500	2025/1/24	gpt-3.5-turbo	39.8	100	4		500	329,381	659	102,815	206	432,196	0.3189
CoT	MATH-500	2025/1/22	Doubao-lite-32k	59	100	4		500	336,370	673	143,571	287	479,941	0.0255
CoT	MATH-500	2025/1/24	gpt-4o	68	100	4		500	329,332	659	223,356	447	552,688	3.0569
CoT	MATH-500	2025/1/22	Qwen2.5-72B-Instruct	80.2	100	4		500	338,549	677	280,466	561	619,015	0.349
CoT	MATH-500	2025/1/24	Llama-3.3-70B-Instruct	71.2	100	4		500	342,879	686	271,342	543	614,221	0.3463
CoT	MATH-500	2025/1/24	Qwen2.5-7B-Instruct	69.8	100	4		500	354,049	708	263,155	526	617,204	0
CoT	MATH-500	2025/1/24	Llama-3.1-8B-Instruct	25.8	100	4		500	342,879	686	282,689	565	625,568	0
CoT	MATH-500	2025/1/24	Internllm2_5-7B	46.6	100	4		500	332,883	666	213,891	428	546,774	0
CoT	MATH-500	2025/1/24	Qwen2-1.5B-Instruct	15.2	100	4		500	349,049	698	187,328	375	536,377	0
CoT	MATH-500	2025/1/24	Qwen2-0.5B-Instruct	6.2	100	4		500	349,049	698	200,139	400	549,188	0
CoT	MATH-500	2025/1/24	deepseek-r1:1.5b	49.4	100	4		500	341,549	683	857,580	1,715	1,199,129	0
PoT	MATH-500	2025/2/10	gpt-3.5-turbo	28.8	83.8	4		500	239,902	480	32,014	64	271,916	0.168
PoT	MATH-500	2025/2/10	Doubao-lite-32k	32.6	68	4		500	254,377	509	48,771	98	303,148	0.0144
PoT	MATH-500	2025/2/10	gpt-4o	46.2	86.4	4		500	241,357	483	99,603	199	340,960	1.5994
PoT	MATH-500	2025/2/10	Qwen2.5-72B-Instruct	47.2	82.2	4		500	242,549	485	170,823	342	413,372	0.233
PoT	MATH-500	2025/2/10	Llama-3.3-70B-Instruct	42.6	80.2	4		500	253,879	508	249,717	499	503,596	0.2839
PoT	MATH-500	2025/2/10	Qwen2.5-7B-Instruct	39.6	74.4	4		500	258,549	517	150,263	301	408,812	0
PoT	MATH-500	2025/2/10	Llama-3.1-8B-Instruct	25.4	68.4	4		500	253,879	508	208,392	417	462,271	0
PoT	MATH-500	2025/2/10	Internllm2_5-7B	15	32.4	4		500	247,883	496	120,826	242	368,709	0
PoT	MATH-500	2025/2/10	Qwen2-1.5B-Instruct	0.8	2.2	4		500	248,509	497	538,361	1,077	786,870	0
PoT	MATH-500	2025/2/10	Qwen2-0.5B-Instruct	0	0	4		500	253,549	507	183,653	367	437,202	0
PoT	MATH-500	2025/2/10	deepseek-r1:1.5b	1	1.6	4		500	245,549	491	785,518	1,571	1,031,067	0
SC-CoT	MATH-500	2025/2/10	gpt-3.5-turbo	40.8	100	4	temperature=1, path_num=5	500	345,411	691	705,408	1,411	1,050,819	1.2308
SC-CoT	MATH-500	2025/2/10	Doubao-lite-32k	65.8	99.8	4	temperature=1, path_num=5	500	362,390	725	715,613	1,431	1,078,003	0.0734
SC-CoT	MATH-500	2025/2/10	gpt-4o	74.6	100	4	temperature=1, path_num=5	500	345,347	691	1,149,778	2,300	1,495,125	12.3611
SC-CoT	MATH-500	2025/2/10	Qwen2.5-72B-Instruct	79.8	100	4	temperature=1, path_num=5	500	1,775,395	3,551	1,506,954	3,014	3,282,349	1.8504
SC-CoT	MATH-500	2025/2/10	Llama-3.3-70B-Instruct	72.4	100	4	temperature=1, path_num=5	500	1,797,045	3,594	1,368,466	2,737	3,165,511	1.7845
SC-CoT	MATH-500	2025/2/10	Qwen2.5-7B-Instruct	71.2	100	4	temperature=1, path_num=5	500	1,855,922	3,712	1,299,553	2,599	3,155,475	0
SC-CoT	MATH-500	2025/2/10	Llama-3.1-8B-Instruct	19.8	99.8	4	temperature=1, path_num=5	500	1,734,545	3,469	1,756,289	3,513	3,490,834	0
SC-CoT	MATH-500	2025/2/10	Internllm2_5-7B	9.2	97.4	4	temperature=1, path_num=5	500	1,994,983	3,990	1,254,893	2,510	3,249,876	0
SC-CoT	MATH-500	2025/2/10	Qwen2-1.5B-Instruct	2	89.4	4	temperature=1, path_num=5	500	1,805,170	3,610	1,333,854	2,668	3,139,024	0
SC-CoT	MATH-500	2025/2/10	Qwen2-0.5B-Instruct	2.2	98.8	4	temperature=1, path_num=5	500	1,808,691	3,617	988,991	1,978	2,797,682	0
SC-CoT	MATH-500	2025/2/10	deepseek-r1:1.5b	46.8	99.2	4	temperature=1, path_num=5	500	1,858,874	3,718	12,109,294	24,219	13,968,168	0
ReAct-Pro*	MATH-500	2025/2/10	gpt-3.5-turbo	23.8	100	4	max_steps=10	500	3,708,461	7,417	124,253	249	3,832,714	2.0406
ReAct-Pro*	MATH-500	2025/2/10	Doubao-lite-32k	47.2	100	4	max_steps=10	500	4,234,620	8,469	154,046	308	4,388,666	0.186
ReAct-Pro*	MATH-500	2025/2/10	gpt-4o	54	100	4	max_steps=10	500	5,834,537	11,669	318,718	637	6,153,255	17.7735
ReAct-Pro*	MATH-500	2025/2/10	Qwen2.5-72B-Instruct	62.8	100	4	max_steps=10	500	5,747,268	11,495	379,849	760	6,127,117	3.4541
ReAct-Pro*	MATH-500	2025/2/10	Llama-3.3-70B-Instruct	64.6	100	4	max_steps=10	500	5,223,611	10,447	418,268	837	5,641,879	3.1806
ReAct-Pro*	MATH-500	2025/2/10	Qwen2.5-7B-Instruct	48.8	100	4	max_steps=10	500	4,646,708	9,293	343,532	687	4,990,240	0
ReAct-Pro*	MATH-500	2025/2/10	Llama-3.1-8B-Instruct	28.8	100	4	max_steps=10	500	7,486,706	14,973	1,276,923	2,554	8,763,629	0
ReAct-Pro*	MATH-500	2025/2/10	Internllm2_5-7B	14.8	100	4	max_steps=10	500	11,831,496	23,663	2,354,609	4,709	14,186,105	0
ReAct-Pro*	MATH-500	2025/2/10	Qwen2-1.5B-Instruct	8.2	100	4	max_steps=10	500	8,430,774	16,862	556,287	1,113	8,987,061	0
ReAct-Pro*	MATH-500	2025/2/10	Qwen2-0.5B-Instruct	0.6	100	4	max_steps=10	500	18,137,392	36,275	1,305,048	2,610	19,442,440	0
ReAct-Pro*	MATH-500	2025/2/10	deepseek-r1:1.5b	24.4	100	4	max_steps=10	500	20,729,970	41,460	9,447,378	18,895	30,177,348	0
ToT	MATH-500	2025/2/10	gpt-3.5-turbo	9.8	100	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	9,711,244	19,422	290,523	581	10,001,767	5.2914
ToT	MATH-500	2025/2/10	Doubao-lite-32k	1.2	94.2	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	5,338,500	10,677	226,000	452	5,564,500	0.2371
ToT	MATH-500	2025/2/10	gpt-4o	3.2	100	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	14,881,985	29,764	360,447	721	15,242,432	40.8094
ToT	MATH-500	2025/2/10	Qwen2.5-72B-Instruct	10.8	100	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	15,657,730	31,315	381,631	763	16,039,361	9.0421
ToT	MATH-500	2025/2/10	Llama-3.3-70B-Instruct	1.4	69.8	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	14,099,500	28,199	570,000	1,140	14,669,500	8.2699
ToT	MATH-500	2025/2/10	Qwen2.5-7B-Instruct	1.4	91.6	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	9,749,000	19,498	418,500	837	10,167,500	0
ToT	MATH-500	2025/2/10	Llama-3.1-8B-Instruct	1.8	90.8	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	7,729,000	15,458	1,306,000	2,612	9,035,000	0
ToT	MATH-500	2025/2/10	Internllm2_5-7B	0.2	99	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	7,515,000	15,030	835,500	1,671	8,350,500	0
ToT	MATH-500	2025/2/10	Qwen2-1.5B-Instruct	0.8	97.2	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	4,408,000	8,816	127,000	254	4,535,000	0
ToT	MATH-500	2025/2/10	Qwen2-0.5B-Instruct	0	96.2	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	5,590,500	11,181	406,000	812	5,996,500	0
ToT	MATH-500	2025/2/10	deepseek-r1:1.5b	0.4	71.6	4	search_type=bfs, b=1, max_depth=6, max_steps=6, generation_n=1, evaluation_n=3, evaluation_type=vote, use_llm_completion=true	500	1,831,000	3,662	110,500	221	1,941,500	0

Default settings:

temperature = 0  (except for SC-CoT)

LLM prices:

• LLM prices:
  • gpt-3.5-turbo:
    • 0.5$/1M tokens (input)
    • 1.5$/1M tokens (output)
  • Doubao-lite-32k (1 USD = 7.3249 CNY):
    • 0.04096$/1M tokens (input)
    • 0.08200$/1M tokens (output)
  • gpt-4o-2024-08-06:
    • 2.50$ /1M input tokens (input)
    • 10$ /1M output tokens (output)
  • Qwen2.5-72B-Instruct and Llama-3.3-70B-Instruct:
    • Prices can be found https://cloud.siliconflow.cn/.
  • Other open source LLMs:
    • Deployed locally, please check the OmAgent repository for more information.
    • Cost is not considered in the leaderboard.

Pass Rate*: The pass rate is calculated by evaluating the percentage of predictions that are valid, where a prediction is valid if it is neither empty nor null.

Performance Comparison of Different Agents and VLMs on MME-RealWorld

Agent	VLMs	Score	Pass Rate	Total Input Tokens	Total Output Tokens	All Tokens
ZoomEye	Qwen2.5-VL-72B-Instruct	51.56	99.81	76,808,965	1,276,460	78,085,425
ZoomEye	Qwen2.5-VL-7B-Instruct	48.06	96.50	94,418,593	1,472,836	95,891,429
IO	Qwen2.5-VL-72B-Instruct	44.47	100.00	6,174,490	2,114	6,176,604
ZoomEye	InternVL2.5-8B	43.42	99.34	153,857,588	2,017,170	155,874,758
IO	InternVL2.5-8B	42.95	100.00	2,779,778	2,335	2,782,113
IO	Qwen2.5-VL-7B-Instruct	42.86	100.00	6,174,490	2,114	6,176,604
ZoomEye	Llava-v1.5-7B	31.60	98.86	113,073,261	1,368,724	114,441,985
IO	Llava-v1.5-7B	24.79	100.00	734,868	17,036	751,904
V*	seal_vqa & seal_vsm	15.14	72.37	-	-	-

Compare to original agent repositories

Algorithm	Dataset	Eval Time	LLM	Framework	Score
CoT	gsm8k	2025/1/7	gpt-3.5-turbo	Original repo	79.23
CoT	gsm8k	2025/1/7	gpt-3.5-turbo	OmAgent	78.70
CoT	AQuA	2025/1/7	gpt-3.5-turbo	Original repo	60.63
CoT	AQuA	2025/1/7	gpt-3.5-turbo	OmAgent	61.02
PoT	gsm8k	2025/1/7	gpt-4o-mini	Original repo	86.35
PoT	gsm8k	2025/1/7	gpt-4o-mini	OmAgent	88.25
ReAct	AQuA	2025/1/7	gpt-3.5-turbo	Original repo	35.04
ReAct	AQuA	2025/1/7	gpt-3.5-turbo	OmAgent	34.25
ReAct	HotpotQA	2025/1/8	gpt-3.5-turbo	Original repo	28.00
ReAct	HotpotQA	2025/1/8	gpt-3.5-turbo	OmAgent	27.40

Note:

• The original repo is the official repository of the agent implementation.
• OmAgent is the implementation of the agent in this project.
• There is no official implementation of SC-CoT.

Comparison ReAct with ReAct-Pro

Algorithm	Dataset	Eval Time	LLM	Score	Pass Rate
ReAct	gsm8k	2025/1/7	gpt-3.5-turbo	38.13	100.00
ReAct-Pro	gsm8k	2025/1/7	gpt-3.5-turbo	74.91	99.39
ReAct	AQuA	2025/1/7	gpt-3.5-turbo	34.25	97.64
ReAct-Pro	AQuA	2025/1/7	gpt-3.5-turbo	64.57	98.03

🔗 Related works

Open Agent Leaderboard is built on top of the OmAgent repository.

🙏 Acknowledgments

We extend our deepest gratitude to the authors and contributors of the following datasets: gsm8k, AQuA, MATH-500, and agent algorithms: CoT, SC-CoT, PoT, ReAct, ToT, and LLMs: gpt-3.5-turbo, Doubao-lite-32k, gpt-4o, Qwen2.5-72B-Instruct, Qwen2.5-7B-Instruct, Qwen2-1.5B-Instruct, Qwen2-0.5B-Instruct, Llama-3.3-70B-Instruct, Llama-3.1-8B-Instruct, Internllm2_5-7B, deepseek-r1:1.5b.

⭐️ Citation

If you find our repository beneficial, please cite our repository:

@misc{open-agent-leaderboard,
    title={Open Agent Leaderboard},
    author={Om AI Lab},
    year={2025},
    publisher={GitHub},
    howpublished={\url{https://github.com/om-ai-lab/open-agent-leaderboard}}
}

🔔 Follow us

You can follow us on X and Discord for more updates and discussions.

🤝 Contributing

Feel free to submit issues and pull requests.

📝 License

This project is licensed under the MIT License.