Qwen3-Coder

Top Builders

Explore the top contributors showcasing the highest number of app submissions within our community.

Qwen3-Coder

Qwen3-Coder is Alibaba Cloud's dedicated coding model, released on July 22, 2025. The flagship variant, Qwen3-Coder-480B-A35B-Instruct, is a mixture-of-experts model trained on 7.5 trillion tokens across 358 programming languages, with 70% of the training data being code. It supports 256K tokens of native context, extensible to 1M tokens with extrapolation, making it suited to repository-scale tasks and multi-file agentic workflows.

General
Release date	22 Jul 2025
Developer	Qwen / Alibaba Cloud
Type	Open-weight coding LLM (MoE)
License	Apache 2.0
GitHub	QwenLM/Qwen3-Coder
Hugging Face	Qwen3-Coder-480B-A35B-Instruct
Documentation	qwenlm.github.io/blog/qwen3-coder

Core Features

480B/35B MoE architecture: 480B total parameters with 35B active per token, using 160 experts with 8 activated per inference step.
256K native context: natively processes 256,000 tokens, with extrapolation support up to 1,000,000 tokens.
358 programming languages: trained on a broad code corpus covering mainstream and niche languages.
Agent RL post-training: long-horizon reinforcement learning trains the model to solve real-world tasks through multi-turn tool interactions.
Apache 2.0: weights are available for commercial use and fine-tuning.

Benchmarks

Benchmark	Score
SWE-Bench Verified	69.6%

SWE-Bench Verified scores are state-of-the-art among open models at release, comparable to Claude Sonnet 4.

Tools and Resources

Qwen3-Coder Blog Post: release announcement with benchmark details.
GitHub (QwenLM/Qwen3-Coder): model code and usage examples.
Hugging Face: model weights.
Qwen Code: open-source terminal agent built on Qwen3-Coder.
Together AI: hosted API access.

Ecosystem and Integrations

Paired with Qwen Code, an open-source terminal coding agent with GitHub Actions support.
Accessible via the Alibaba Cloud DashScope API using an OpenAI-compatible endpoint.
Available on Together AI, LM Studio, and Ollama for local and cloud inference.

To use Qwen3-Coder via API, get an API key on the Qwen API Platform. For local agentic coding, see the Qwen Code terminal agent.

Edit on GitHub

Qwen Qwen3-Coder AI technology Hackathon projects

Discover innovative solutions crafted with Qwen Qwen3-Coder AI technology, developed by our community members during our engaging hackathons.

AuraOS

Aura is a sovereign, neuro-symbolic AI operating system designed to make advanced artificial intelligence more transparent, efficient, collaborative, and accountable. Instead of sending every instruction directly to a language model, Aura can locally interpret natural-language intent, organize it into structured semantic and routing representations, identify relevant code or knowledge, and prepare a bounded task before model use is required. Aura combines deterministic routing, a 4,096-primitive lexical system, polysynthetic six-slot intent structures, visual code topology, guarded human-agent workflows, and a Learning Arena where verified experiences can be tested before becoming reusable knowledge. Its Observatory allows users to see how Aura understood an instruction, while the Human Agent Arena supports node-specific dialogue, approval gates, verification, and preserved attempt artifacts—including failed work that may help future refactoring. Aura is built around human authority. Models can propose, explain, investigate, and assist, but they cannot silently expand scope, bypass evidence requirements, or merge changes without approval. The result is an AI architecture focused not only on capability, but on sovereignty, inspectability, learning, and responsible action.

Probably runs on Samsung S25

Hybrid Router is an intelligent AI inference system that dynamically selects the best execution path for each user query instead of relying on a single large language model. The system first analyzes the incoming prompt to identify the task type, estimate its complexity, and determine whether it can be solved deterministically or requires generative AI. For structured tasks such as mathematical calculations, JSON validation, regular expression verification, and date/time operations, the router invokes specialized deterministic tools that produce fast, accurate, and reproducible results without consuming LLM tokens. For more complex natural language and coding tasks, the router attempts local inference using OpenVINO-optimized models running on Intel hardware, reducing latency and API costs. If the local model is unlikely to provide a sufficiently reliable answer or the task exceeds its capabilities, the system automatically falls back to Fireworks AI models for high-quality remote inference. The routing decisions are driven by task classification, confidence estimation, and configurable thresholds, allowing the system to balance accuracy, response time, and operational cost. The architecture is modular, making it easy to add new tools, local models, or routing strategies in the future. The project also includes benchmarking and evaluation components that measure routing accuracy, latency, model utilization, and fallback frequency to continuously improve routing performance. By combining deterministic tools, local inference, and cloud-based language models into a single adaptive pipeline, Hybrid Router delivers efficient, scalable, and cost-aware AI inference while maintaining high response quality across a wide range of tasks.

EduSense AI

The motivation for this project stems from the observation that many children experience anxiety due to bullying, classroom trauma, academic pressure, or social challenges, yet these symptoms often go unnoticed until they become severe. The platform combines behavioral questionnaires, emotional assessments, classroom observations, and parental feedback to generate a comprehensive anxiety profile for each child. Machine learning models analyze multiple behavioral and emotional indicators to estimate anxiety risk levels, while Large Language Models (LLMs) provide personalized explanations and evidence-based recommendations tailored to parents, teachers, school counselors, and psychologists.

Audisor-amd-track1

This project is a Track 1 benchmark agent designed for the AMD Developer Hackathon Act II. It reads the provided /input/tasks.json, classifies each task, applies deterministic local tools when correctness is provable, and falls back to Fireworks AI only when a task requires model inference or local validation cannot safely accept the result. The runtime enforces the contest contract by writing a valid /output/results.json with exact task_id and answer fields. Remote calls use only FIREWORKS_BASE_URL, FIREWORKS_API_KEY, and models from ALLOWED_MODELS at runtime. The goal is to pass the accuracy gate while keeping recorded Fireworks token usage low.

ISEEKAIGO

Most AI companions today are built around static prompts layered on top of large language models. They can generate impressive responses, but they rarely evolve beyond the conversation in front of them. ISEEKAIGO is building the next generation of AI companionship by combining generative AI with persistent memory, human-inspired needs, autonomous behavior, and evolving cognition. Instead of simply responding, our AI learns from experience, self-learning, adapts over time, and develops its own behavioral context, making every interaction more personal and meaningful. Our vision extends beyond creating a better chatbot. We are building a foundation for Human-like AI that can interact with people over long periods while becoming increasingly capable through experience. By breaking this vision into scalable milestones, we aim to transform AI companionship into a platform for future autonomous AI applications such as building physical-AI core. We believe this project represents a meaningful step toward the future of Human-like AI, combining ambitious research with the ability to ship real products and bring advanced AI experiences to millions of users.

EventFlow – Parallel AI Vendor Negotiation

EventFlow automates the procurement of event vendors in Pakistan's $50 billion wedding industry, which lacks structured tools. In this market, deals are typically negotiated via phone calls and WhatsApp, with no APIs, no pricing data, and no way to compare offers. To use EventFlow, customers input their total budget, specify event categories, and set a range for negotiation flexibility within a Flutter app. An Analyzer Agent (powered by a large language model) then optimally distributes the budget across the selected categories. A Matchmaker service calculates dynamic, category-specific pricing—such as per-head costs for catering, guest-threshold scaling for decorators, and flat rates for photographers and DJs. Instead of negotiating with each vendor one at a time, EventFlow activates a separate Negotiation Agent for each matched vendor, allowing all negotiations to run in parallel and in real-time within the app. Every negotiation utilizes Fireworks AI's Kimi K2 model, running on AMD’s MI300X. This setup enables fast and cost-effective handling of five to eight simultaneous negotiations per event, avoiding the delays associated with a sequential queue. On the backend, FastAPI and PostgreSQL serve as the authoritative source for every change in negotiation state. A state synchronization service mirrors every write operation to Firestore in real-time, ensuring that both customer and vendor applications remain in sync. Trust is enforced on the server-side: each counter-offer is validated against a floor price established during the matchmaking process and is never accepted from the client side. Any attempts to bypass this will result in an instant HTTP 400 error. Finally, an Aggregator Agent synthesizes the best combination of finalized deals across all vendor categories into one confirmed booking.