# The Bitter Lesson of Diffusion Language Models for Agentic Workflows: A Comprehensive Reality Check arXiv: https://arxiv.org/abs/2601.12979 · HTML: https://mlsanity.com/paper/2601.12979/ ## Key facts - Cited by 6 papers in the map (data through 14 Oct 2026). - Placed in Discrete and text diffusion, one of 14 directions in Diffusion models. - Builds on 33 of its 49 references that are in the map. - First posted on arXiv 19 Jan 2026. ## Placed in - Diffusion models / [Discrete and text diffusion](https://mlsanity.com/diffusion/discrete-and-text-diffusion/) - LLM agents and tool use / [General LLM agent frameworks and benchmarks](https://mlsanity.com/agents/general-llm-agent-frameworks-and-benchmarks/) ## Builds on (most cited first, 30 of 33) - [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023) - [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022) - [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023) - [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023) - [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023) - [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023) - [A Survey on Large Language Model based Autonomous Agents](https://mlsanity.com/paper/2308.11432/) (2023) - [τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains](https://mlsanity.com/paper/2406.12045/) (2024) - [ALFWorld: Aligning Text and Embodied Environments for Interactive Learning](https://mlsanity.com/paper/2010.03768/) (2020) - [Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding](https://mlsanity.com/paper/2505.22618/) (2025) - [AgentBoard: An Analytical Evaluation Board of Multi-turn LLM Agents](https://mlsanity.com/paper/2401.13178/) (2024) - [ScienceWorld: Is your Agent Smarter than a 5th Grader?](https://mlsanity.com/paper/2203.07540/) (2022) - [SegDiff: Image Segmentation with Diffusion Probabilistic Models](https://mlsanity.com/paper/2112.00390/) (2021) - [BabyAI: A Platform to Study the Sample Efficiency of Grounded Language Learning](https://mlsanity.com/paper/1810.08272/) (2018) - [Diffusion Probabilistic Modeling for Video Generation](https://mlsanity.com/paper/2203.09481/) (2022) - [ToolACE: Winning the Points of LLM Function Calling](https://mlsanity.com/paper/2409.00920/) (2024) - [Large Language Model Agent: A Survey on Methodology, Applications and Challenges](https://mlsanity.com/paper/2503.21460/) (2025) - [SwiftSage: A Generative Agent with Fast and Slow Thinking for Complex Interactive Tasks](https://mlsanity.com/paper/2305.17390/) (2023) - [Fast-dLLM v2: Efficient Block-Diffusion LLM](https://mlsanity.com/paper/2509.26328/) (2025) - [Autonomous Evaluation and Refinement of Digital Agents](https://mlsanity.com/paper/2404.06474/) (2024) - [Diffusion LLMs Can Do Faster-Than-AR Inference via Discrete Diffusion Forcing](https://mlsanity.com/paper/2508.09192/) (2025) - [Accelerating Diffusion LLMs via Adaptive Parallel Decoding](https://mlsanity.com/paper/2506.00413/) (2025) - [SEAgent: Self-Evolving Computer Use Agent with Autonomous Learning from Experience](https://mlsanity.com/paper/2508.04700/) (2025) - [StateFlow: Enhancing LLM Task-Solving through State-Driven Workflows](https://mlsanity.com/paper/2403.11322/) (2024) - [Understanding and Improving Lexical Choice in Non-Autoregressive Translation](https://mlsanity.com/paper/2012.14583/) (2020) - [AgentRefine: Enhancing Agent Generalization through Refinement Tuning](https://mlsanity.com/paper/2501.01702/) (2025) - [SlotRefine: A Fast Non-Autoregressive Model for Joint Intent Detection and Slot Filling](https://mlsanity.com/paper/2010.02693/) (2020) - [Tool-Planner: Task Planning with Clusters across Multiple Tools](https://mlsanity.com/paper/2406.03807/) (2024) - [Diffusion LLM with Native Variable Generation Lengths: Let [EOS] Lead the Way](https://mlsanity.com/paper/2510.24605/) (2025) - [PreAct: Prediction Enhances Agent's Planning Ability](https://mlsanity.com/paper/2402.11534/) (2024) ## Built on by (most cited first, 6 of 6) - [DLLM Agent: See Farther, Run Faster](https://mlsanity.com/paper/2602.07451/) (2026) - [Continuous Interaction Diffusion: A Diffusion-Native Runtime for Asynchronous Tool-Augmented Reasoning](https://mlsanity.com/paper/2608.10438/) (2026) - [CanvasAnneal: Curriculum Reinforcement Learning for Diffusion Language Models](https://mlsanity.com/paper/2609.13060/) (2026) - [One Readout, Many Repairs: Diffusion-Guided Hierarchical Search for Tool-Agent Repair](https://mlsanity.com/paper/2609.34879/) (2026) - [Does This Action Still Explain the Task? Reverse Scoring for Diffusion Language Model Agents](https://mlsanity.com/paper/2609.38536/) (2026) - [Plan-and-Patch: Diffusion Language Models for Agentic Planning](https://mlsanity.com/paper/2610.10786/) (2026) ## Abstract (arXiv) The pursuit of real-time agentic interaction has driven interest in Diffusion-based Large Language Models (dLLMs) as alternatives to auto-regressive backbones, promising to break the sequential latency bottleneck. However, does such efficiency gains translate into effective agentic behavior? In this work, we present a comprehensive evaluation of dLLMs (e.g., LLaDA, Dream) across two distinct agentic paradigms: Embodied Agents (requiring long-horizon planning) and Tool-Calling Agents (requiring precise formatting). Contrary to the efficiency hype, our results on Agentboard and BFCL reveal a "bitter lesson": current dLLMs fail to serve as reliable agentic backbones, frequently leading to systematically failure. (1) In Embodied settings, dLLMs suffer repeated attempts, failing to branch under temporal feedback. (2) In Tool-Calling settings, dLLMs fail to maintain symbolic precision (e.g. strict JSON schemas) under diffusion noise. To assess the potential of dLLMs in agentic workflows, we introduce DiffuAgent, a multi-agent evaluation framework that integrates dLLMs as plug-and-play cognitive cores. Our analysis shows that dLLMs are effective in non-causal roles (e.g., memory summarization and tool selection) but require the incorporation of causal, precise, and logically grounded reasoning mechanisms into the denoising process to be viable for agentic tasks.