Free API integrations for OpenClaw/NemoClaw using GLM-5, Kimi K2.5, MiniMax, NVIDIA API Catalog, and Kilo CLI. Extracted from AI Code King tutorial (bdNf-KieKTY).
nvapi-XXXX keylibrechat.yaml or env):
providers:
nvidia:
apiKey: "nvapi-XXXX"
baseURL: "https://integrate.api.nvidia.com/v1"
models:
- meta/llama-3.1-70b-instruct
- nvidia/nemotron-70b-instruct
- mistralai/mixtral-8x22b-instruct-v0.1
Boomer_Ang specialist execution (code generation tasks)Base URL: https://open.bigmodel.cn/api/paas/v4
Model: glm-5
NTNTN intent normalization (strong reasoning), Review_Hone validationBase URL: https://api.moonshot.cn/v1
Model: kimi-k2.5
Guide_Ang responses, simple chat completionsBase URL: https://api.minimax.chat/v1
Model: MiniMax-M2.5
npm install -g @kilocode/cli
kilo
/connect # Configure AI provider credentials
Boomer_Ang execution backend for code tasksBoomer_Ang| Free Provider | Best ACHEEVY Role | Task Type | Context Window |
|---|---|---|---|
| NVIDIA NIM (Llama/Nemotron) | Boomer_Ang | Code gen, general | 128k |
| GLM-5 (Zhipu) | Boomer_Ang | Agentic SWE, coding | 137k |
| Kimi K2.5 (Moonshot) | NTNTN, Review_Hone | Reasoning, multimodal | 256k |
| MiniMax M2.5 | Guide_Ang | Fast chat, simple tasks | 197k |
| Kilo CLI | Boomer_Ang (code backend) | Code execution | Per-model |
| OpenCode | Boomer_Ang (code backend) | Code execution | Per-model |
When estimatedCostTokens needs to be $0 (free tier):
When cost budget exists: