Urgent.News

What's breaking now, across thousands of outlets.

AI

I Run Claude Code on $0/Month: A Practical Guide to Free LLM APIs

I Run Claude Code on $0/Month: A Practical Guide to Free LLM APIs AI coding assistants are incredible — until the API bill arrives. If you've ever watched your Claude Code or Cursor usage tick upward and wondered whether there's a cheaper way, this guide is for you. The good news: in 2026, the free tier landscape for LLM APIs is genuinely good. Not "free trial for 7 days" good — permanently free,…

AI coding assistants can be expensive, but Claude Code provides a way to use free large language model (LLM) APIs. Starting in 2026, the free tier options for LLM APIs are actually quite good, with permanent free access and no need for a credit card. However, free tiers and their details change frequently, so it's important to stay updated. This guide provides a practical setup that the author actually uses.

The most valuable free providers include Google AI Studio with Gemini 2.5 Flash, Groq with Llama 3.3 70B, NVIDIA NIM offering over 100 open models, OpenRouter with 35+ free models, Cerebras with Llama 3.3 70B, and Mistral with Mistral Small/Medium models. A complete comparison of rate limits, context windows, and credit card requirements is maintained at freellm.net, a directory of 489 free models across 30 providers.

To get started, simply obtain a free API key from one of these providers. Google AI Studio and Groq are recommended for coding due to their model quality and speed. Sign up for the provider, create an API key, and copy it. With the key in hand, you can begin using Claude Code as normal by setting two environment variables. For Google AI Studio, set ANTHROPIC_BASE_URL and ANTHROPIC_AUTH_TOKEN with your Google AI Studio key. For Groq, set ANTHROPIC_BASE_URL to https://api.groq.com/openai/v1.

Before committing your key to daily workflows, test it using freellm.net's playground, which allows you to paste a key and chat with any model directly in the browser without any installation. Keep in mind that free tiers come with limits, such as Google AI Studio's 500 requests/day on Flash. While these limits may be sufficient for personal coding, they may not be enough for a team. Additionally, note that there is no permanent free tier for Claude, and the quality of non-Anthropic models varies.

The author uses Google AI Studio's Gemini 2.5 Flash via Google AI Studio's free tier for their AI-assisted coding, while Groq is used when speed is needed. If a rate limit is hit, the author switches providers, which only takes about 30 seconds due to the simplicity of changing environment variables. For students, indie hackers, or those who are cost-conscious, there's no reason to pay for API access when these free tiers are available. The entire setup takes less time than reading this article to understand.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

Move a Claude Code task to a teammate without sharing a login

A coding task often stalls because the person who wrote the brief cannot run it right now. Passing a provider login to someone else blurs who executed the work and what access they had.

  • Handoff file includes task description, outcome, scope, and starting details
  • Executor uses own account with Claude Code access, no shared login
  • Reviewer checks changed files, commit/PR URL, test results, and visual check

Running Qwen 3.8 Flash Next (125B) on a RTX 4090 – 100 T/s on a Desktop

1. What was released / announced Niko1221’s Strata repo shows that the new Qwen 3.8 Flash Next (125B) model can be run on a consumer‑grade RTX 4090 at an impressive 100 trillion tokens per second…

  • Qwen 3.8 Flash Next (125B) model runs on RTX 4090 at 100 T/s
  • Model quantized to 4-bits with torch.compile and custom CUDA kernel
  • Enables on-premises inference, eliminating cloud token fees and reducing latency

The 'Terminator scenario': How realistic is the AI apocalypse?

OpenAI boss Sam Altman and Anthropic chief Dario Amodei warn of the risks of overly rapid AI development. More experts now think AI could wipe out humanity. How realistic is that?

  • Skynet AI system breaks free in 2029, leading to nuclear war and autonomous killer robots.
  • Artificial intelligence expert Reinhard Karger warns of unintended catastrophic consequences.

More from Monday 5 October →