Nvidia has free kimi K3, glm 5.3 and deepseek V4.1 flash APIs.
Stop collecting AI bookmarks and use one. All three appear on NVIDIA Build with free endpoints. The point is to prototype: send a task, let the AI generate code or an answer, then see whether it deserves a place in
[Breaking news] “NVIDIA Build free inference frame” where you can try Kimi K3 and DeepSeek is a hot topic!
! NVIDIA's developer portal (K3, DeepSeek V4.1 Flash, GLM 5.3 and other notable models are available as Free Endpoint (prototype API)) ・OpenAI compatible Base
The board picked up fresh options
GPT-6 Luna, Gemini Omni 1.1 Flash, GLM 5.3 all FREE: Claude Code: $100 on Pro, $250 on Max in cloud credit if you already have an active sub, claim by oct 7: NVIDIA NIM: GLM 5.3, GLM 5.3 Flash, DeepSeek V4.1 Flash,
I asked Opus 5.5 nd DeepSeek V4.1 Flash to "Make the Boston Dynamics Dog".
See the results.
You have reached the end of the archive
All of deepseek