mlsanityData to 7 Oct 2026

Qwen3-Coder-Next Technical Report

arXiv 2603.00729 ↗First posted 28 Feb 2026cs.CLcited by 60

On the map: LLM agents and tool use → Software engineering coding agents

Abstract (arXiv)

We present Qwen3-Coder-Next, an open-weight language model specialized for coding agents. Qwen3-Coder-Next is an 80-billion-parameter model that activates only 3 billion parameters during inference, enabling strong coding capability with efficient inference. In this work, we explore how far strong training recipes can push the capability limits of models with small parameter footprints. To achieve this, we perform agentic training through large-scale synthesis of verifiable coding tasks paired with executable environments, allowing learning directly from environment feedback via mid-training and reinforcement learning. Across agent-centric benchmarks including SWE-Bench and Terminal-Bench, Qwen3-Coder-Next achieves competitive performance relative to its active parameter count. We release both base and instruction-tuned open-weight versions to support research and real-world coding agent development.

Key facts. Cited by 60 papers in the map (data through 7 Oct 2026). Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use. Builds on 42 of its 57 references that are in the map. First posted on arXiv 28 Feb 2026.

Builds on

42 of 57 references.

Closest ancestors · 7

Same direction, newest first

  1. 2026MegaFlow: Large-Scale Distributed Orchestration System for the Agentic Era12 Jan
  2. 2025SWE-RM: Execution-free Feedback For Software Engineering Agents26 Dec
  3. 2025SWE-rebench: An Automated Pipeline for Task Collection and Decontaminated Evaluation of Software Engineering Agents26 May
  4. 2024Training Software Engineering Agents and Verifiers with SWE-Gym30 Dec
  5. 2024OpenHands: An Open Platform for AI Software Developers as Generalist Agents23 Jul
  6. 2024Agentless: Demystifying LLM-based Software Engineering Agents1 Jul
  7. 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering6 May

Other references · 35

  1. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  2. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,193
  3. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,748
  4. 2022Survey of Hallucination in Natural Language Generationcited by 2,056
  5. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Codecited by 1,906
  6. 2019Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelismcited by 1,829
  7. 2024MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmarkcited by 1,758
  8. 2023Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generationcited by 1,332
  9. 2025gpt-oss-120b & gpt-oss-20b Model Cardcited by 1,258
  10. 2024Qwen2.5-Coder Technical Reportcited by 1,151
  11. 2018Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Taskcited by 1,120
  12. 2025DeepSeek-V3.2: Pushing the Frontier of Open Large Language Modelscited by 755
23 more
  1. 2023Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLscited by 667
  2. 2025Kimi K2: Open Agentic Intelligencecited by 523
  3. 2026Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfacescited by 515
  4. 2024StarCoder 2 and The Stack v2: The Next Generationcited by 484
  5. 2024CRUXEval: A Benchmark for Code Reasoning, Understanding and Executioncited by 294
  6. 2025MathArena: Evaluating LLMs on Uncontaminated Math Competitionscited by 275
  7. 2024ToolACE: Winning the Points of LLM Function Callingcited by 204
  8. 2025APIGen-MT: Agentic Pipeline for Multi-Turn Data Generation via Simulated Agent-Human Interplaycited by 158
  9. 2024Vulnerability Detection with Code Language Models: How Far Are We?cited by 133
  10. 2025CWM: An Open-Weights LLM for Research on Code Generation with World Modelscited by 87
  11. 2025Agent Learning via Early Experiencecited by 79
  12. 2024CTIBench: A Benchmark for Evaluating LLMs in Cyber Threat Intelligencecited by 61
  13. 2024FullStack Bench: Evaluating LLMs as Full Stack Coderscited by 56
  14. 2025CWEval: Outcome-driven Evaluation on Functionality and Security of LLM Code Generationcited by 55
  15. 2024Fewer Truncations Improve Language Modelingcited by 33
  16. 2025OJBench: A Competition Level Code Benchmark For Large Language Modelscited by 22
  17. 2025SWE-Flow: Synthesizing Software Engineering Data in a Test-Driven Mannercited by 18
  18. 2026SWE-Universe: Scale Real-World Verifiable Environments to Millionscited by 16
  19. 2025PlotCraft: Pushing the Limits of LLMs for Complex and Interactive Data Visualizationcited by 11
  20. 2024JumpCoder: Go Beyond Autoregressive Coder via Online Modificationcited by 7
  21. 2026From Completion to Editing: Unlocking Context-Aware Code Infilling via Search-and-Replace Instruction Tuningcited by 5
  22. 2025AthenaBench: A Dynamic Benchmark for Evaluating LLMs in Cyber Threat Intelligencecited by 4
  23. 2026Evaluating and Achieving Controllable Code Completion in Code LLMcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 60.

60 papers.

Sort:

By area:

  1. 2026OpenClaw-RL: Train Any Agent Simply by Talkingcited by 83
  2. 2026SpecBench: Measuring Reward Hacking in Long-Horizon Coding Agentscited by 25
  3. 2026CodeTracer: Towards Traceable Agent Statescited by 11
  4. 2026REAM: Merging Improves Pruning of Experts in LLMscited by 9
  5. 2026Orchard: An Open-Source Agentic Modeling Frameworkcited by 7
  6. 2026CLI-Anything: Towards Agent-Native Computer Usecited by 7
  7. 2026Group-Graph Policy Optimization for Long-Horizon Agentic Reinforcement Learningcited by 7
  8. 2026The Verification Horizon: No Silver Bullet for Coding Agent Rewardscited by 6
  9. 2026PM-Bench: Evaluating Prospective Memory in LLM Agentscited by 6
  10. 2026OScaR: The Occam's Razor for Extreme KV Cache Quantization in LLMs and Beyondcited by 5
  11. 2026Rethinking the Role of Efficient Attention in Hybrid Architecturescited by 5
  12. 2026SWE-Bench ProMax: Benchmarking Agents on Large-Scale Multilingual Code Refactoringcited by 5
38 more
  1. 2026WebTestBench: Evaluating Computer-Use Agents towards End-to-End Automated Web Testingcited by 4
  2. 2026DeonticBench: A Benchmark for Reasoning over Rulescited by 4
  3. 2026Revisiting DAgger in the Era of LLM-Agentscited by 4
  4. 2026boldsymbolf-OPD: Stabilizing Long-Horizon On-Policy Distillation with Freshness-Aware Controlcited by 4
  5. 2026PEEK: Context Map as an Orientation Cache for Long-Context LLM Agentscited by 4
  6. 2026Masking Stale Observations Helps Search Agents -- Until It Doesn't: A Regime Map and Its Mechanismcited by 4
  7. 2026DiscoLoop: Looping Discrete Embeddings and Continuous Hidden States for Multi-hop Reasoningcited by 4
  8. 2026Mach-Mind-4-Flash Technical Reportcited by 4
  9. 2026Memory Transfer Learning: How Memories are Transferred Across Domains in Coding Agentscited by 3
  10. 2026AdaExplore: Failure-Driven Adaptation and Diversity-Preserving Search for Efficient Kernel Generationcited by 3
  11. 2026EIBench: A Simulator-Based Benchmark and Turn-Credit RL for Emotion Managementcited by 3
  12. 2026TDAD: Test-Driven Agentic Development - Reducing Code Regressions in AI Coding Agents via Graph-Based Impact Analysiscited by 2
  13. 2026AssertLLM2: A Comprehensive LLM Benchmark for Assertion Generation from Design Specificationscited by 2
  14. 2026Morphing into Hybrid Attention Modelscited by 2
  15. 2026BlockPilot: Instance-Adaptive Policy Learning for Diffusion-based Speculative Decodingcited by 2
  16. 2026Stage-Aware Communication Scheduling for Disaggregated LLM Servingcited by 1
  17. 2026Terminus-4B: Can a Smaller Model Replace Frontier LLMs at Agentic Execution Tasks?cited by 1
  18. 2026CopT: Contrastive On-Policy Thinking with Continuous Spaces for General and Agentic Reasoningcited by 1
  19. 2026Teaching LLMs to Self-Evolve: Cultivating Core Meta-Skills with Reinforcement Learningcited by 1
  20. 2026Learning from 53.6K Real-World Developer Edits of AI-Generated Codecited by 1
  21. 2026RISE-RL: Rubric-Informed Selective Exploration for Open-Ended Reinforcement Learningcited by 1
  22. 2026Massive Activations in Hybrid Linear Attention Large Language Models: Pre-Attention Spikes and Inter-Spike Plateauscited by 1
  23. 2026EVOMAL: Self-Poisoning in Self-Evolving Coding Agentscited by 1
  24. 2026NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harnesscited by 1
  25. 2026An Empirical Study of Harness Design for Coding Agentscited by 1
  26. 2026CRANE: Constrained Reasoning Injection for Code Agents via Nullspace Editingcited by 0
  27. 2026DASH: Fast Differentiable Architecture Search for Hybrid Attention in Minutes on a Single GPUcited by 0
  28. 2026Agentic Separation Logic Specification Synthesiscited by 0
  29. 2026Code as a Weapon: A Consensus-Labeled Prompt Bank for Measuring Coding-Model Compliance with Malicious-Code Requestscited by 0
  30. 2026Frontier LLM-based agents can overcome the ontology curation bottleneck for natural phenotypescited by 0
  31. 2026LLM Evolution as an Industry-Scale Ecosystem: A Lifecycle Perspective on Continual Learningcited by 0
  32. 2026Code evolution for link prediction in complex networkscited by 0
  33. 2026An Empirical Study of Security Calibration in Large Language Models for Codecited by 0
  34. 2026Minos: A Multi-Agent Collaborative Framework for Provenance-Based Backward Trackingcited by 0
  35. 2026Enhancing SLMs for Sustainable Code Optimization in Radio-Astronomycited by 0
  36. 2026Consilience for Verifier-Free Test-Time Scalingcited by 0
  37. 2026No One Model Catches Every Harm: Benchmarking Content Moderation Across Safety Scenarioscited by 0
  38. 2026NoTB: Oracle-Free Triage of LLM-Generated RTL via Cross-Model Formal Consensuscited by 0

50 most cited of 60 All 60 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026Coaching Qwen3 Coder 30B to Think Like a CodeClash Arena Agent
  2. 2025A Self-Improving Coding Agent
  3. 2025MapCoder-Lite: Distilling Multi-Agent Coding into a Single Small LLM
  4. 2026Steer, Don't Solve: Training Small Critic Models for Large Code Agents
  5. 2026CodeMidas: Scaling Agentic Coding RL Environments from Code Itself
  6. 2026Cross-Benchmark Transfer from RL on Agentic Coding Tasks
  7. 2026Scaling Agentic Verifier for Competitive Coding
  8. 2026KAT-Coder-V2.5 Technical Report