Qwen3 Coder 30B A3B Instruct
Qwen3 Coder 30B A3B Instruct is Alibaba's coding-focused MoE model with native 256K context, repository-scale understanding, and structured tool calling.
Model Overview
Capabilities, design details, and architectural traits
Qwen3 Coder 30B A3B Instruct - Coding-Focused Mixture-of-Experts Model
Qwen3 Coder 30B A3B Instruct is Qwen's instruction-tuned coding model built specifically for software development workflows. Rather than serving as a general-purpose language model, it is positioned as a coding-focused Mixture-of-Experts (MoE) model designed for code generation, repository-scale understanding, and integration with coding tools.
Its defining characteristic is the combination of an MoE architecture, native 256K-token context window, and a coding-oriented function calling format intended for development environments.
| Documented characteristic | Description |
|---|---|
| Mixture-of-Experts architecture | Activates only a subset of experts during inference while using a larger expert pool. |
| Native long context | Supports a native 256K-token context window, with documentation describing extension to 1M tokens using YaRN. |
| Repository-scale understanding | Long-context design is optimized for understanding large codebases and repositories. |
| Coding-oriented function calling | Uses a function call format designed for coding workflows and tool integration. |
| Coding platform support | Documented to support coding environments including Qwen Code and CLINE. |
Instead of emphasizing conversational applications, the model is designed around development tasks where long-context code comprehension and structured tool interaction are central. Its documented workflow targets scenarios such as navigating large repositories, generating code, and interacting with coding platforms through its dedicated function-calling interface.
Benchmark Performance
Independent evaluations · Artificial Analysis
Accuracy & Capability Details
Compare Models Side-by-Side
Evaluate specifications, pricing, and independent benchmark indices
| Model Details | |||
|---|---|---|---|
| General Info | |||
| Provider | Alibaba | Anthropic | Anthropic |
| Release Date | July 31, 2025 | September 22, 2026 | September 28, 2026 |
| Knowledge Cutoff | Jun 2025 | - | Jun 2026 |
| Context & Limits | |||
| Context Window | 262K | 1M Best Context Window | 1M Best Context Window |
| Pricing (per 1M tokens) | |||
| Input Pricing | $0.15 Best Input Pricing | $4 | $2 |
| Output Pricing | $0.75 Best Output Pricing | $20 | $10 |
| Modalities | |||
| Inputs | text | textimagefile | textimagefile |
| Outputs | text | text | text |
| Benchmarks (0-100) | |||
| Intelligence Index | 9.6 | 57.6 Best Intelligence Index | 56.0 |
| Coding Index | - | - | - |
| Agentic Index | - | - | - |
Humanity's Last Exam
Extremely difficult logical reasoning and knowledge.
Long Context Reasoning
Logical reasoning over long context windows.
SciCode Benchmark
Scientific coding and mathematical modeling.
Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.
Explore more from Alibaba
Other models by Alibaba
Qwen3.8 Max
Qwen3.8 2.4T A95B
Qwen3.8-Flash-Next
Qwen3.8 27B
Top AI Models
Leading alternatives by intelligence score
