# Liquid AI 8B-A1B MoE

Hacker Approved page: https://hackerapproved.com/p/liquid-ai-8b-a1b-moe/

8 people recommended Liquid AI 8B-A1B MoE in Hacker News comments, 4 had mixed feelings and 7 criticized it, November 2025 to October 2026. It was recommended in 5 of 12 months, in 7 threads. It does not meet the Hacker Approved badge rule.

4 of the 8 people who recommend it commented in one thread, “Liquid AI reveals 8B-A1B MoE trained on 38T”.

**Verdict:** People recommend it for fast local inference, including on CPUs and integrated GPUs. The most common caveats are that it is not very smart, weak at world knowledge, and below expectations.

Summary written by Claude from 21 comments.

## Why people recommend it

- Very fast local inference (8 comments in the sample): "Bad at translation, at least to Russian. Very fast though, about 2x faster than Gemma 4 e2b on my CPU." ([source](https://news.ycombinator.com/item?id=48334150)) "it's my go-to model for whenever i need fast local inference" ([source](https://news.ycombinator.com/item?id=48746877)) "I liked LFM2-8B-A1B for its speed on cpu or integrated gpu" ([source](https://news.ycombinator.com/item?id=47984754))
- Useful for focused, single-purpose tasks (3 comments in the sample): "I’ve also found LFM2.5 8B surprisingly useful for single-focus tasks" ([source](https://news.ycombinator.com/item?id=49417855)) "the Liquid AI models are already very good at a lot of this" ([source](https://news.ycombinator.com/item?id=50283444)) "it executed it extremely well" ([source](https://news.ycombinator.com/item?id=48328208))

## What people warn about

- Not very smart, limited knowledge (3 comments in the sample): "it's expectedly not favorable for world knowledge" ([source](https://news.ycombinator.com/item?id=48336249)) "these things clearly don't know much yet" ([source](https://news.ycombinator.com/item?id=48330775)) "It's not very smart, to say the least." ([source](https://news.ycombinator.com/item?id=49418519))
- Below expectations on performance (3 comments in the sample): "I just assumed it would perform better, given all the advancements in the space." ([source](https://news.ycombinator.com/item?id=48339114)) "I was never able to get anything useful out of them" ([source](https://news.ycombinator.com/item?id=47518380)) "It did not perform as well as I expected." ([source](https://news.ycombinator.com/item?id=48330346))

## Other small language models people recommend

- [OpenAI Luna](https://hackerapproved.com/p/luna/): 171 recommend, 31 mixed, 46 criticize
- [Claude Haiku](https://hackerapproved.com/p/claude-haiku/): 147 recommend, 48 mixed, 113 criticize
- [Apple Foundation Models](https://hackerapproved.com/p/apple-foundation-models/): 24 recommend, 17 mixed, 31 criticize
- [Ministral](https://hackerapproved.com/p/ministral/): 14 recommend, 3 mixed, 5 criticize

## The badge

A product earns the Hacker Approved badge when at least 10 people recommend it, recommenders outnumber critics at least 2 to 1 (mixed feelings count toward neither side), it was recommended in at least 4 of the 12 months and in at least 5 different threads, and no single thread supplies more than half of the people who recommend it. Health and money products are never badged.

## Cite as

Hacker Approved, "Liquid AI 8B-A1B MoE", https://hackerapproved.com/p/liquid-ai-8b-a1b-moe/ (8 recommend, 4 mixed, 7 criticize; Hacker News comments, November 2025 to October 2026).

Counts updated 3 Nov 2026, with data through October 31, 2026.