Yesterday, we dropped our latest AI model: Gemini 3.7 Flash ⚡️ our most intelligent workhorse model yet for coding and agents.
Just 1 day later, we’re already impressed by the creativity we’ve seen.
Here are a few of our favorite ways we’ve seen people use 3.7 Flash so far ⬇️
Yesterday, we released Gemini 3.7 Flash — our most intelligent workhorse model yet for coding and agents.
See what developers, users and our partners have to say about Gemini 3.7 Flash ⬇️
old-coder swaps code review for an evidence gauntlet. Human approves the test plan, agent does TDD, then mutation testing, coverage, property tests. You review metrics, not diffs. MIT. You stopped reading agent diffs months ago. Now it's a methodology.
Qwen 3.8 - 27B MLX-Serve 4bit model, made the best version of the famous Pagoda test so far for me. Using @pidotdev as the coding agent, and MLX-Serve as the backend.
Yesterday we introduced Gemini 3.7 Flash, our most intelligent workhorse model yet for coding and agents.
Watch @Ammaar, Member of the Technical Staff at @GoogleDeepMind, use 3.7 Flash to turn simple prompts into playable games 👾
We promised open weights for Qwen3.8. Now, time to meet them! 🎉
⚡ Qwen3.8-27B:
- A native multimodal dense model. With just 27B parameters, it outperforms Qwen3.7-Plus overall and shines in real-world coding & office workflows.
- 262K native context, easily extendable to 1M tokens via YaRN.
- Built for builders. Highly efficient, high-quality, and licensed under Apache 2.0.
🚀 The open weights for Qwen3.8-2.4T-A95B (Max-level) have also been released recently.
Whether you're shipping lightweight applications with Qwen3.8-27B locally or building agents with Qwen3.8-2.4T-A95B, they're yours now!
Download, deploy, and build something we haven't imagined yet. 👀👇
- Hugging Face:
- ModelScope: