A100 80GB GPUs in Google Colab via Google AI Pro plans
Melvin Vivas · X post · 2026-10-10 · Open on X
Topics: Fine-tuning & Model Customization, AI Dev Tools & Productivity, LLM Fundamentals · Level: intermediate
Summary
Google AI Pro subscriptions now bundle premium Google Colab benefits, including A100 GPUs with 80GB VRAM, and Ultra subscribers get H100s. That headroom lets you run and fine-tune mid-sized open models like Gemma 4 right in the browser.
Key points
- Google AI Pro plans include premium Colab benefits with A100 GPUs (80GB VRAM).
- Google AI Ultra subscribers can also use H100 GPUs in Colab.
- 80GB VRAM is enough to run Gemma 4 31B in bf16 for inference.
- You can do a full fine-tune of models up to Gemma 4 E4B in bf16.
- LoRA tuning works on even larger models (the quoted post is cut off before giving the size).
- No local GPU setup is needed. It all runs in the browser through Colab notebooks.
Resources mentioned
- Google Colab · tool · colab.research.google.com · free · recommended by both Bashiri Smith & Melvin Vivas
Free GPU notebooks.
Also in: Run LightOnOCR-3 0.8B and Other Open Models on a Free Colab T4 (Melvin Vivas on X · notes), Try Liquid AI's open d1-3B decision model in Colab (Melvin Vivas on X · notes), Google Colab notebooks for new open models, built by a coding agent (Melvin Vivas on X · notes), Liquid AI's LFM2.5-24B-A2B: a local MoE model for agent tasks (Melvin Vivas on X · notes) and 27 more - Google AI Pro · tool · one.google.com · paid
Paid Google AI subscription plan that now includes premium Colab benefits. - Google AI Ultra · tool · one.google.com · paid
Google's paid AI subscription plan that gives access to its top Gemini models.
Also in: Gemini 4 Argon Announced (Google AI Ultra Only) (Melvin Vivas on X · notes) - Gemma 4 · tool · ai.google.dev · free
Google's family of open-weight models in several sizes, built to run on devices and offline, with multimodal and agentic abilities, and open to fine-tuning.
Also in: Meetily: Open-Source Local AI Meeting Notes App (Melvin Vivas on X · notes), New Gemma Developer Documentation Site (Melvin Vivas on X · notes), A day of work done from a phone with Grok Bot agents (Melvin Vivas on X · notes), Agent tasks: Unsloth fine-tuning, ComfyUI face swap, script writing (Melvin Vivas on X · notes) and 32 more
Try this
- Try running or fine-tuning a Gemma 4 model on a Colab A100 if you have a Google AI Pro plan.
- Full fine-tune Gemma 4 E4B in bf16 on Colab A100 80GB.
- LoRA-tune a larger Gemma 4 model in Colab.
More in Fine-tuning & Model Customization
- Fine-tuning Meta's Muse Glimmer on Kalshi Prediction Outcomes
- Fine-tuned Qwen3.5-9B beat frontier models for Coinbase fraud task
- The Open Env Arena: Agents Design RL Environments, Ranked on a Leaderboard
- Fine-tune and Run Local Models for Free on Google Colab's T4 GPU
- Free Colab notebooks for fine-tuning and running local models