A Systematic Survey of LLM-Based Agentic AI Frameworks for Multi-Agent Coordination and Interoperability
DOI:
https://doi.org/10.66279/y29vex64Keywords:
Agentic AI, Multi-Agent Systems, Agent Communication Protocols, Agent Orchestration, Autonomous AgentsAbstract
Large Language Models (LLMs) have spurred the development of agentic artificial intelligence systems that can reason autonomously, plan, use tools, integrate memory, and carry out multi-step tasks. Unlike conventional prompt-response systems, LLM-based agents extend generative models to goal-oriented architectures that can decompose complex objectives, interact with external environments, and coordinate actions in iterative workflows. The survey systematically reviews LLM-based agent frameworks, multi-agent collaboration architectures, internal communication mechanisms, emerging cross-system interoperability protocols, and open research challenges. The review uses a PRISMA-based methodology, including literature from 2020 to March 2026, with particular focus on prominent frameworks such as LangChain, AutoGPT, AutoGen, MetaGPT, CAMEL, ChatDev, and CrewAI. The survey discusses the transition from modular chain-oriented to stateful graph-oriented and autonomous execution models and presents a systematic taxonomy of multi-agent collaboration architectures, including hierarchical, peer-to-peer, and role-based models. It also reviews the main internal communication mechanisms used to facilitate coordination and information sharing between LLM-based agents and distinguishes them from emerging interoperability protocols designed to support interaction across heterogeneous agents, tools, services, and frameworks. The analysis indicates that the promise of LLM-based agents for scalable automation, collaborative reasoning, and complex workflow execution comes with significant challenges in long-horizon reliability, evaluation standardization, communication security, cost-efficient orchestration, governance, and the interpretability of emergent multi-agent behavior. The survey covers architectural evolution, collaboration patterns, communication mechanisms, framework-level characteristics, and open research challenges. This provides a structured foundation for future research on reliable and trustworthy agentic artificial intelligence systems.
Downloads
References
[1] V. I. Slyusar, Z. Gomolka, and Y. P. Kondratenko, “General Characteristics of the Large Language Models and Comparative Analysis,” AI in Education Systems: Successful Cases and Perspectives, pp. 1–20, Oct. 2025, doi: 10.1201/9788743809258-1/GENERAL-CHARACTERISTICS-LARGE-LANGUAGE-MODELS-COMPARATIVE-ANALYSIS-SLYUSAR-GOMOLKA-KONDRATENKO. DOI: https://doi.org/10.1201/9788743809258-1
[2] T. Ma and X. Chen, “Emerging Trends in LLM Agent Development,” Jan. 2026, doi: 10.2139/SSRN.6119146. DOI: https://doi.org/10.2139/ssrn.6119146
[3] M. Abou Ali, F. Dornaika, and J. Charafeddine, “Agentic AI: a comprehensive survey of architectures, applications, and future directions,” Artificial Intelligence Review 2025 59:1, vol. 59, no. 1, pp. 11-, Nov. 2025, doi: 10.1007/S10462-025-11422-4. DOI: https://doi.org/10.1007/s10462-025-11422-4
[4] Y. Mendoza Juan, “Development of a multi-agent, LLM-driven system to enhance human-machine interaction: integrating DSPy with modular agentic strategies and logical reasoning layers for the autonomous generation of smart contracts,” Jul. 15, 2024, Universitat Politècnica de Catalunya. Accessed: May 18, 2026. [Online]. Available: https://hdl.handle.net/2117/414255
[5] E. Davis, “Building Custom AI Workflows Using LangChain Tools,” ThinkTide Global Research Journal, vol. 5, no. 4, pp. 54–62, Nov. 2024, Accessed: May 18, 2026. [Online]. Available: http://thinktidejournal.com/index.php/TGRJ/article/view/53
[6] W. Cugunov, Unlocking the Power of Auto-GPT and Its Plugins: Implement, Customize, and Optimize Auto-GPT for Building Robust AI Applications. Packt Publishing, Sep. 13, 2024.
[7] E. Oliveira, K. Fischer, and O. Stepankova, “Multi-agent systems: which research for which applications,” Rob. Auton. Syst., vol. 27, no. 1–2, pp. 91–106, Apr. 1999, doi: 10.1016/S0921-8890(98)00085-2. DOI: https://doi.org/10.1016/S0921-8890(98)00085-2
[8] U. Eswaran, V. Eswaran, V. Eswaran, and K. Murali, “Agentic AI in Space Exploration: Autonomous Decision-Making Beyond Earth,” pp. 111–126, 2025, doi: 10.1007/978-3-031-89424-4_8. DOI: https://doi.org/10.1007/978-3-031-89424-4_8
[9] S. H. Shaikh, “LLM-Based Multi-agent Systems: Frameworks, Evaluation, Open Challenges, and Research Frontiers,” Communications in Computer and Information Science, vol. 2827 CCIS, pp. 149–170, 2026, doi: 10.1007/978-3-032-15632-7_9/SAVE-RESEARCH. DOI: https://doi.org/10.1007/978-3-032-15632-7_9
[10] T. Guo et al., “Large Language Model based Multi-Agents: A Survey of Progress and Challenges,” IJCAI International Joint Conference on Artificial Intelligence, pp. 8048–8057, Jan. 2024, Accessed: Aug. 27, 2026. [Online]. Available: https://arxiv.org/pdf/2402.01680
[11] J. Li et al., “Fundamental Capabilities and Applications of Large Language Models: A Survey,” ACM Comput. Surv., vol. 58, no. 2, Sep. 2025, doi: 10.1145/3735632. DOI: https://doi.org/10.1145/3735632
[12] J. Schneider, “Agentic AI Prompt Engineering: Advancing Generative AI Patterns Conceptually,” pp. 1–8, Jan. 2026, doi: 10.1109/ICITDA68167.2025.11332344. DOI: https://doi.org/10.1109/ICITDA68167.2025.11332344
[13] SapkotaRanjan, ShresthaRashik, RijalMadhav, and KarkeeManoj, “LangChain vs. LangGraph vs. LangSmith: Taxonomies of Agentic AI Toolchains for End-to-End Orchestration,” Sep. 2025, doi: 10.36227/TECHRXIV.175695645.52670060/V1. DOI: https://doi.org/10.36227/techrxiv.175695645.52670060/v1
[14] M. Fırat and S. Kuleli, “What if GPT4 Became Autonomous: The Auto-GPT Project and Use Cases,” Journal of Emerging Computer Technologies, vol. 3, no. 1, pp. 1–6, Mar. 2024, doi: 10.57020/JECT.1297961. DOI: https://doi.org/10.57020/ject.1297961
[15] Q. Wu et al., “AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation,” Aug. 2023, Accessed: Aug. 10, 2026. [Online]. Available: https://arxiv.org/pdf/2308.08155
[16] [S. Hong et al., “MetaGPT: Meta Programming for a Multi-Agent Collaborative Framework,” in Proc. 12th Int. Conf. Learn. Representations (ICLR), 2024.
[17] Guohao Li∗ HasanAbedAlKaderHammoud* HaniItani* DmitriiKhizbullin, “CAMEL:Communicative Agents for ‘Mind’ Exploration of Large Language Model Society”, Accessed: May 22, 2026. [Online]. Available: https://arxiv.org/pdf/2303.17760
[18] G. I. Rashed, A. O. M. Bahageel, H. A. I. Gony, A. Badjan, and H. I. Shaheen, “AI-Driven Multi-Agent Demand Response Framework for Residential Load Optimization Using CrewAi,” 2025 IEEE 20th Conference on Industrial Electronics and Applications, ICIEA 2025, 2025, doi: 10.1109/ICIEA65512.2025.11149013. DOI: https://doi.org/10.1109/ICIEA65512.2025.11149013
[19] Y. Lei et al., “Large Language Model Agents: A Comprehensive Survey on Architectures, Capabilities, and Applications,” Dec. 2025, doi: 10.20944/PREPRINTS202512.2119.V1. DOI: https://doi.org/10.20944/preprints202512.2119.v1
[20] A. A. Jafari, C. Ozcinar, and G. Anbarjafari, “A Lightweight Modular Framework for Constructing Autonomous Agents Driven by Large Language Models: Design, Implementation, and Applications in AgentForge,” Jan. 2026, Accessed: May 20, 2026. [Online]. Available: https://arxiv.org/pdf/2601.13383
[21] O. Topsakal and T. C. Akinci, “Creating Large Language Model Applications Utilizing LangChain: A Primer on Developing LLM Apps Fast,” International Conference on Applied Engineering and Natural Sciences, vol. 1, no. 1, pp. 1050–1056, Jul. 2023, doi: 10.59287/ICAENS.1127. DOI: https://doi.org/10.59287/icaens.1127
[22] L. Kuligin, J. Zaldívar, and M. Tschochohei, Generative AI on Google Cloud with LangChain: Design Scalable Generative AI Solutions with Python, LangChain, and Vertex AI on Google Cloud. Packt Publishing, 2024.
[23] D. Tolis, J. Rytilahti, O. Weerakoon, P. Puhtila, E. Kaila, and T. Mäkilä, “Democratizing AI Development: A Feature-Based Categorization of API Platforms, Development Frameworks, LCNC and AIaaS Platforms for LLM-Based Applications,” Journal of Software Engineering Research and Development, vol. 14, no. 1, pp. 62–87, May 2026, doi: 10.5753/JSERD.2026.5969. DOI: https://doi.org/10.5753/jserd.2026.5969
[24] M. Isham, A. Raj, K. A. Shetty, and K. Adamsab, “Branch LLM: A Local Branching Memory Interface for Exploratory Conversations,” pp. 572–577, Dec. 2025, doi: 10.1109/INSPIRE67328.2025.11300653. DOI: https://doi.org/10.1109/INSPIRE67328.2025.11300653
[25] C. Z. Chang Yang†, Z. Y. Zijin Hong, and J. S. Ninghao Liu, “Graph-based Agent Memory: Taxonomy, Techniques, and Applications”, Accessed: May 22, 2026. [Online]. Available: https://arxiv.org/pdf/2602.05665
[26] O. Elizarov, “Architecture of Applications Powered by Large Language Models,” 2024, Accessed: May 20, 2026. [Online]. Available: http://www.theseus.fi/handle/10024/868581
[27] Z. Lei, “LangGraph vs LangChain: Which Framework Is Best for AI Agents?,” MetaVision Journal of Multidisciplinary Studies (MVJMS), vol. 1, no. 4, pp. 16–26, Nov. 2024, Accessed: May 20, 2026. [Online]. Available: https://sinojournals.com/index.php/mvjms/article/view/10
[28] L. Reddy Alva, B. Pandey, L. R. Alva, and B. Pandey, “Agentic AI systems in the age of generative models: architectures, cloud scalability, and real-world applications,” Artificial Intelligence Review 2025 59:3, vol. 59, no. 3, pp. 88-, Jan. 2026, doi: 10.1007/S10462-025-11458-6. DOI: https://doi.org/10.1007/s10462-025-11458-6
[29] M. Leo, F. Tan, T. Miao, and G. Anand, “From threat to trust: assessing security risks of agentic AI systems,” International Journal of Information Security 2026 25:1, vol. 25, no. 1, pp. 23-, Jan. 2026, doi: 10.1007/S10207-025-01185-Y. DOI: https://doi.org/10.1007/s10207-025-01185-y
[30] D. S et al., “AGENTIC AI: AUTONOMOUS INTELLIGENT AGENTS USING LARGE LANGUAGE MODELS AND CLOUD COMPUTING FOR REAL-WORLD TASK AUTOMATION,” JETIR, vol. 13, no. 5, pp. d5–d9, 2026, Accessed: May 20, 2026. [Online]. Available: https://www.jetir.org/view?paper=JETIR2605302
[31] L. Lin and Y. Liu, “Multimodal Large Models Toward AGI,” Multimodal Large Models, pp. 263–336, 2026, doi: 10.1007/978-981-95-4929-0_5. DOI: https://doi.org/10.1007/978-981-95-4929-0_5
[32] “GitHub - Significant-Gravitas/AutoGPT: AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters. · GitHub.” Accessed: Aug. 27, 2026. [Online]. Available: https://github.com/Significant-Gravitas/AutoGPT?utm_source=chatgpt.com
[33] A. Bandi, B. Kongari, R. Naguru, S. Pasnoor, and S. V. Vilipala, “The Rise of Agentic AI: A Review of Definitions, Frameworks, Architectures, Applications, Evaluation Metrics, and Challenges,” Future Internet 2025, Vol. 17, Page 404, vol. 17, no. 9, p. 404, Sep. 2025, doi: 10.3390/FI17090404. DOI: https://doi.org/10.3390/fi17090404
[34] Q. Meng et al., “Agent Harness for Large Language Model Agents: A Survey,” Apr. 2026, doi: 10.20944/PREPRINTS202604.0428.V3. DOI: https://doi.org/10.20944/preprints202604.0428.v1
[35] S. Brady, “Springdrift: An Auditable Persistent Runtime for LLM Agents with Case-Based Memory, Normative Safety, and Ambient Self-Perception,” Apr. 2026, Accessed: May 21, 2026. [Online]. Available: https://arxiv.org/pdf/2604.04660
[36] L. Wang et al., “A survey on large language model based autonomous agents,” Frontiers of Computer Science 2024 18:6, vol. 18, no. 6, pp. 186345-, Mar. 2024, doi: 10.1007/S11704-024-40231-1. DOI: https://doi.org/10.1007/s11704-024-40231-1
[37] B. Zhang et al., “Breaking Agents: Compromising Autonomous LLM Agents Through Malfunction Amplification,” EMNLP 2025 - 2025 Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference, pp. 34964–34976, 2025, doi: 10.18653/V1/2025.EMNLP-MAIN.1771. DOI: https://doi.org/10.18653/v1/2025.emnlp-main.1771
[38] D. Goyal and A. Gautam, “Introduction to LangChain Framework,” Textual Intelligence Large Language Models and Their Real-World Applications., pp. 253–285, Jan. 2025, doi: 10.1002/9781394287499.CH11. DOI: https://doi.org/10.1002/9781394287499.ch11
[39] Z. Feng et al., “Multi-agent embodied AI: advances and future directions,” Science China Information Sciences 2026 69:5, vol. 69, no. 5, pp. 151202-, Mar. 2026, doi: 10.1007/S11432-025-4820-4. DOI: https://doi.org/10.1007/s11432-025-4820-4
[40] Z. Deng et al., “Multi-Agent Systems and Social Intelligence: A Comprehensive Survey on Collaboration, Communication, and Emergent Behaviors,” Dec. 2025, doi: 10.2139/SSRN.5957495. DOI: https://doi.org/10.2139/ssrn.5957495
[41] Z. Zhang, Y. Zhang, D. Zeng, K. Liu, and J. Zhao, “Beware of the Woozle Effect: Exploring and Mitigating Hallucination Propagation in Multi-Agent Debate,” IEEE Trans. Audio Speech Lang. Process., 2026, doi: 10.1109/TASLPRO.2026.3675803. DOI: https://doi.org/10.1109/TASLPRO.2026.3675803
[42] S. Sarkar, “Enterprise Usage of AI Agents: A Comprehensive Survey on Applications, Frameworks, and Future Productivity Impact,” Mar. 2026, doi: 10.2139/SSRN.6405998. DOI: https://doi.org/10.2139/ssrn.6405998
[43] Q. Wang et al., “MegaAgent: A Large-Scale Autonomous LLM-based Multi-Agent System Without Predefined SOPs,” pp. 4998–5036, Accessed: May 23, 2026. [Online]. Available: https://github.com/ DOI: https://doi.org/10.18653/v1/2025.findings-acl.259
[44] C. Qian et al., “ChatDev: Communicative Agents for Software Development,” vol. 1, pp. 15174–15186, Accessed: May 23, 2026. [Online]. Available: https://github.com/OpenBMB/ChatDev. DOI: https://doi.org/10.18653/v1/2024.acl-long.810
[45] N. Krishnan, “Advancing Multi-Agent Systems Through Model Context Protocol: Architecture, Implementation, and Applications,” 2025.
[46] C. Li et al., “ACPS: Agent Collaboration Protocols for the Internet of Agents,” Proceedings of 2025 9th IEEE International Conference on Network Intelligence and Digital Content, IC-NIDC 2025, pp. 342–346, 2025, doi: 10.1109/IC-NIDC67200.2025.11390349. DOI: https://doi.org/10.1109/IC-NIDC67200.2025.11390349
[47] T. Kehkashan et al., “From benchmarks to deployment: a comprehensive review of agentic AI evaluation,” Artificial Intelligence Review 2026 59:8, vol. 59, no. 8, pp. 167-, Apr. 2026, doi: 10.1007/S10462-026-11571-0. DOI: https://doi.org/10.1007/s10462-026-11571-0
[48] L. Geng and E. Y. Chang, “REALM-Bench: A Benchmark for Evaluating Multi-Agent Systems on Real-world, Dynamic Planning and Scheduling Tasks,” Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, vol. 1-A, pp. 2652–2662, Apr. 2026, doi: 10.1145/3770854.3785692;SUBPAGE:STRING:BASIC. DOI: https://doi.org/10.1145/3770854.3785692
[49] Y. Wang et al., “Internet of Agents: Fundamentals, Applications, and Challenges,” IEEE Trans. Cogn. Commun. Netw., vol. 12, pp. 4476–4501, 2026, doi: 10.1109/TCCN.2025.3623369. DOI: https://doi.org/10.1109/TCCN.2025.3623369
[50] K. Jiang et al., “KoMA: Knowledge-Driven Multi-Agent Framework for Autonomous Driving with Large Language Models,” IEEE Transactions on Intelligent Vehicles, vol. 10, no. 10, pp. 4655–4668, 2025, doi: 10.1109/TIV.2024.3488793. DOI: https://doi.org/10.1109/TIV.2024.3488793
Downloads
Published
Data Availability Statement
NA
Issue
Section
Categories
License
Copyright (c) 2026 Journal of Smart Algorithms and Applications (JSAA)
This work is licensed under a Creative Commons Attribution 4.0 International License.
Journal of Smart Algorithms and Applications (JSAA) content is published under a Creative Commons Attribution 4.0 International (CC BY 4.0) License. This means that content is freely available to all readers upon publication, and content is published as soon as production is complete.
Journal of Smart Algorithms and Applications (JSAA) seeks to publish the most influential papers that will significantly advance scientific understanding. Selected articles must present new and widely significant data, syntheses, or concepts. They should merit recognition by the wider scientific community and the general public through publication in a reputable scientific journal.
