Skip to main content
Back to Pulse
TechCrunch

Tiny startup Arcee AI built a 400B-parameter open source LLM from scratch to best Meta’s Llama

Read the full articleTiny startup Arcee AI built a 400B-parameter open source LLM from scratch to best Meta’s Llama on TechCrunch

What Happened

30-person startup Arcee AI has released a 400B model called Trinity, which it says is one of the biggest open source foundation models from a U.S. company.

Our Take

Okay, 30 people building a 400B model is legitimately wild. That's not a toy — if Trinity actually competes with Llama 3, they've cracked something on efficiency or training cost. Here's the thing though: open-source foundation models are getting commoditized fast. Arcee's talented, sure, but there's no moat in training the same architecture everyone else is.

The upside is someone big (Meta, Hugging Face, maybe a cloud provider) acquires them. The downside is they're out of runway in 18 months with no revenue model. Neither's great for using Trinity in production.

What To Do

Benchmark Trinity against Llama 3 on your specific use case before it becomes vapor.

Cited By

React

Loading comments...