Free Claude Code

Nemo Code by ClawdWorks

Want these models inside your existing Claude Code install instead? Keep your memories, `CLAUDE.md`, MCP servers, hooks, and local workflow while swapping model backends with [AnyModel](https://kevster527.gumroad.com/l/anymodel).

Free AI coding agent. Zero subscription. One command.

Nemo Code gives you a Claude Code-style CLI experience - tools, file editing, shell commands, MCP servers, autocompact-style workflows, and optional Telegram access - powered by open/free model backends such as NVIDIA-hosted OSS models.

Built around:

• Claude Code CLI as a runtime dependency
• LiteLLM as the local Anthropic-compatible proxy
• NVIDIA NIM hosted model endpoints

Nemo Code does not bundle or redistribute Claude Code.

Quick Start

Mac / Linux / Git Bash

curl -fsSL https://raw.githubusercontent.com/kevdogg102396-afk/free-claude-code/master/install.sh | bash

Windows PowerShell

irm https://raw.githubusercontent.com/kevdogg102396-afk/free-claude-code/master/install.ps1 | iex

Then open a new terminal and run:

clawdworks

2026 Model Slots

Nemo maps Claude Code's model slots to current NVIDIA NIM model IDs. Switch in the TUI with `/model`.

Claude Code slot	NVIDIA model ID	Best for
Sonnet (default)	moonshotai/kimi-k2.6	Current top coding/default model
Opus	qwen/qwen3.5-397b-a17b	Large reasoning and complex tasks
Haiku	minimaxai/minimax-m2.7	Fast agentic coding and document work

Additional selectable install-time models:

• z-ai/glm5.1
• nvidia/nemotron-3-super-120b-a12b
• openai/gpt-oss-120b

Model availability can change on NVIDIA's side. If a model disappears or rate limits heavily, rerun `clawdworks models` or set `NEMO_MODEL=`.

`NEMO_MODEL` controls the Sonnet/default route. Opus and Haiku stay mappe