Stele now runs natively inside the new DeepSeek Harness.
Agents can use Stele knowledge directly while they work, with referenced knowledge highlighted inline and full session context available from the UI. You can see exactly what Stele tools were used, what knowledge the
DeepSeek Harness went from nothing to 150,000 GitHub stars in days, it's MIT licensed, and the plugin system is insane.
You can rip out the core, swap the model for Claude or OpenAI, and build your own tooling. Worth a look if Claude Code is your daily driver.
Keep your setup. Add Claude Opus 5, GPT-5.6, Gemini 3.7 Flash, Grok 4.6, Kimi K3, GLM 5.3, DeepSeek V4 Pro, Qwen3.8 Max and more leading models.
3/ the API is included too
This is the part developers will probably find more interesting. The access isn’t restricted to chat interface. Developers can create an API Key and make live calls to DeepSeek-V4-Flash without first recharging the account.
You have reached the end of the archive
All of deepseek