Run Claude Code on free models

What this is

A free, step-by-step template from Taha's reels — the exact tools, prompts and links, laid out so you can follow along and ship. Comment the keyword on the reel and it also lands in your DMs.

Hey! 👋 Here's what I promised you — with an honest warning first:

Can't pay for a Claude subscription? You can keep the Claude Code terminal you know and plug other, free models into it. You are not getting Claude itself: you get models like DeepSeek, Kimi or GLM inside the Claude Code interface. Great for small projects and for learning — not the same quality as Claude.

⚠️ free-claude-code is an unofficial, third-party proxy. Its own README says it is "Not affiliated with or endorsed by Anthropic". It sits between Claude Code and the model, and it holds your API keys — read the code before you trust it. It can break when Claude Code updates, and each provider's terms still apply to traffic you send through it. "Free" depends on third-party free tiers that can change or end at any time.

1. free-claude-code — the proxy: Claude Code thinks it is talking to Claude, but the answers come from the provider you pick in its Admin UI.

🔗GitHubhttps://github.com/Alishahryar1/free-claude-code

Install (Windows PowerShell):

& ([scriptblock]::Create((irm "https://raw.githubusercontent.com/Alishahryar1/free-claude-code/main/scripts/install.ps1")))

Install (macOS/Linux):

curl -fsSL "https://raw.githubusercontent.com/Alishahryar1/free-claude-code/main/scripts/install.sh" | sh

Open Free Claude Code (on Linux: fcc-server), add a provider key in the Admin UI, then start Claude Code through it:

fcc-claude

Then pick where the free models come from

2. NVIDIA NIM — create a free API key with an NVIDIA account and use open models like DeepSeek or GLM on NVIDIA's free inference endpoints (rate-limited, and the model list changes).

🔗NVIDIA buildhttps://build.nvidia.com 🔗API keyshttps://build.nvidia.com/settings/api-keys

3. OpenRouter — same idea: one key, and a list of models marked (free) at $0 — plus many more models if you add credits later.

🔗OpenRouterhttps://openrouter.ai 🔗Free models listhttps://openrouter.ai/models?max_price=0

4. Ollama — if you have a PC with a good GPU, run a small model locally: nothing to pay, and your data never leaves your PC. Ollama can also start Claude Code on a local model by itself, no proxy needed.

🔗Ollamahttps://ollama.com 🔗GitHubhttps://github.com/ollama/ollama 🔗Ollama + Claude Codehttps://docs.ollama.com/integrations/claude-code

Install Ollama (Windows):

irm https://ollama.com/install.ps1 | iex

Install Ollama (macOS/Linux):

curl -fsSL https://ollama.com/install.sh | sh

Then launch Claude Code on your local model:

ollama launch claude

If you can afford it later, Claude Pro/Max is still the real thing — free models are for small projects and learning. 🔀

Want more? Go further in the Atlas Room Skool class

This page gives you the tools. The Skool class takes you step by step — full courses and guides to build and sell your own AI app or SaaS, in Darija, with a new lesson every week.

Join the Skool class →

New to vibe coding? See what you'll learn →

🎯LEADS

LTX🎬