ByteDance Open Source
@ByteDanceOSS
Announcing new releases and the latest information about ByteDance open source projects.
Exciting news — Coze is now open source! 🚀 Check it out and show some love. ‘Coze Officially Embrace Open Source ’ Contributions and stars welcome! ⭐️github.com/coze-dev/coze-… ⭐️github.com/coze-dev/cozel… 🔧 #OpenSource #AI #coze
Coze Officially Embrace Open Source github.com/coze-dev/coze-… github.com/coze-dev/coze-… #coze #OpenSource
Seed-X 🔥 a suite of multilingual translation models released by @ByteDanceOSS huggingface.co/collections/By… ✨ instruction/reinforcement learning/reward model ✨ Supports 28 languages, bidirectional translation ✨ Optimized for deployment & inference: 7B with mistral architecture…
🚀 Rspack v1.4.8 is out! We've implemented a Rust watcher to replace the original watchpack package, making HMR ~10% faster. Combined with lazy compilation for lightning-fast dev experience ⚡
Congrats to @cognition_labs! Looking forward to having both foreground & background coding agents working together. In fact Devin is already the top 20 committer to @verl_project . A great assistant for creating well-tested and documented code.
Cognition has signed a definitive agreement to acquire Windsurf. The acquisition includes Windsurf’s IP, product, trademark and brand, and strong business. Above all, it includes Windsurf’s world-class people, whom we’re privileged to welcome to our team. We are also honoring…
🚀 Exciting news for #GoLang & #AI enthusiasts! Meet DeerGo – the Go-powered version of Deerflow, our deep research AI agent! github.com/cloudwego/eino… #Golang #AIAgent #DeepResearch #OpenSource #CloudWeGo
Tar 🔥 Unified multimodal LLM for vision understanding & generation by @ByteDanceOSS huggingface.co/collections/By… ✨ 1.5B/7B - Apache 2.0 ✨ Fully discrete, LLM-aligned visual tokens (TA-Tok) ✨ AR & diffusion detokenizers for high-fidelity generation
Open-source, SWE-bench #1 Agent, transparent pricing usage for the IDE. If you are looking for AI IDEs don't miss @Trae_ai 👉 Check it out: github.com/bytedance/trae… 🔗 SWE-bench leaderboard: swebench.com

If you're in Singapore on 7/11, do not miss this meetup! Talks from the verl community: - LLMs to optimize code performance on real-world repos & verl project updates @sivil_taram - Long-horizon LLM agent training with verl-agent @langfengq Link: lu.ma/e498qhsi
We’ve open-sourced Trae-Agent. You can all `git clone` `cd trae-agent` now github.com/bytedance/TRAE…
SeedEdit: Fast and High-Quality Generative Image Editing @ByteDanceOSS ❤️ Performance: • Usability rate: 56.1% (SeedEdit 1.6: 38.4%, GPT-4o: 37.1%). • Speed: 15 seconds (GPT-4o: 50 seconds). Model Architecture: •High-resolution (1024x1024) image processing based on…
Introducing Agent TARS Beta — a brand new and more powerful Agent TARS! - Agent TARS CLI - Browser Agent driven by Seed-1.5-VL - Native Streaming - Multimodal-friendly Web UI - Layered Agent architecture Blog: agent-tars.com/beta Quick start: agent-tars.com/quick-start
🚀 Join us at the SF AIBrix & vLLM Meetup on June 18th at AWS SF GenAI Loft! Learn from experts at ByteDance, AWS Neuron, and EKS. Discover AIBrix: a scalable, cost-effective control plane for vLLM. Talks, Q&A, pizza, and networking! 🍕🤝 lu.ma/ab2id296
1. BAGEL by @ByteDanceOSS Is an open-source foundation model trained on diverse interleaved multimodal data, outperforming peers in reasoning, manipulation, and understanding arxiv.org/abs/2505.14683 Project Page: bagel-ai.org
A new recipe for training multimodal models 👉 Mixed together various data types: text next to images, video frames after captions, then webpages, etc. This way the model learns to connect what it reads with what it sees. ByteDance proposed and implemented this idea in their…
✨ AIBrix v0.3.0 is here! Cloud-native toolkit for scalable, cost-efficient LLM inference. Features: ✅ KVCache offloading ✅ Adaptive routing ✅ Prod-ready benchmarking Tackles token reuse, low-latency traffic & real-world workload evaluation. #LLM #AIInfrastructure #AIBrix
AIBrix v0.3.0 is officially released! Also checkout our plans for v0.4.0! We ❤️ your feedback! Blog: aibrix.github.io/posts/2025-05-…
Join the SF vLLM Meetup! 🚀 June 18, 2025 @ AWS GenAI Loft (525 Market St). Bytedance AIBrix + AWS Neuron/EKS speakers. Optimize LLMs, connect with devs. RSVP now! #vLLM #GenAI #AIBrix lu.ma/ab2id296
🚀 BAGEL — the Unified Multimodal Model with emergent capabilities and production-ready performance — is finally live! Dive in here: 👉 bagel-ai.org
Very efficient yet powerful 8B coding model trained with @verl_project. Check it out! bytedance-seed-coder.github.io
🚀 Thrilled to introduce Seed-Coder, our new open-source family of 🔥powerful, 🔍transparent, ⚡parameter-efficient code models at 8B scale! 💥 Small model, big results! 🥇 BigCodeBench, FullStack Bench, MHPP: impressive results among lightweight open-source models, including…