Qwen3-Coder-Next Technical Report
This work addresses the need for efficient coding agents in AI research and real-world development, though it is incremental as it builds on existing training recipes and benchmarks.
The paper tackles the problem of developing efficient coding agents by introducing Qwen3-Coder-Next, an 80-billion-parameter model that activates only 3 billion parameters during inference, achieving competitive performance on benchmarks like SWE-Bench and Terminal-Bench relative to its active parameter count.
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.