NVIDIA gives free year-long access to 140+ AI models including GLM 5.2 and MiniMax M3
Roger_M_Taylor · x · 2026-09-06
NVIDIA is offering a free one-year API key that unlocks 140+ models, including GLM 5.2, MiniMax M3, Nemotron-3.5-lightning-30b-a3b, and Meta's Muse-glimmer-30b.
Setup is straightforward: register on build.nvidia.com (phone verification required) to get an nvapi- key, add it in Hermes agent under Settings → Model Provider → Custom with the provided baseurl, then switch between models like z-ai/glm-5.2 or minimaxai/minimax-m3.
For agent builders paying $50-200/month for API access, this effectively zeros out inference costs for agent loops.
More from coding & agent
- GPT-6 Astra mixes Computer Use and direct file writing to build LEGO models — dkundel · 2026-09-06
- Merge Agent Handler adds six callable connectors including TinyFish and Microsoft Graph Security — shensi · 2026-09-06
- Spec-to-APK entirely on an Android phone via Termux CLI and AI — Digital_Otorongo · 2026-09-06
- Astra review: executes hour-long agentic tasks while iterating on the plan without losing the thread — ivan_bezdomny · 2026-09-06
- Stop picking coding models on vibes: run your own A/B/C refactor tests — robleclerc · 2026-09-06
- Orca, the 62k-star ADE for running parallel coding agents, recommended for Mac Mini dev setups — kieranklaassen · 2026-09-06