# Segment Anything Model 3

Hacker Approved page: https://hackerapproved.com/p/meta-segment-anything-model-3/

38 people recommended Segment Anything Model 3 in Hacker News comments, 14 had mixed feelings and 8 criticized it, November 2025 to October 2026. It was recommended in 9 of 12 months, in 22 threads. It meets the Hacker Approved badge rule.

**Verdict:** People recommend it for fast, accurate zero-shot segmentation, streaming video support, and as a teacher model for training smaller ones. Critics say it fits specialised work such as engineering drawings and medical imaging poorly.

Summary written by Claude from 46 comments.

## Why people recommend it

- Seen as a breakthrough in segmentation (4 comments in the sample): "It's really, really good. This feels like a seminal moment for computer vision." ([source](https://news.ycombinator.com/item?id=45984797)) "Segment Anything literaly leapfroged progress on image segmentation." ([source](https://news.ycombinator.com/item?id=46107310)) "Segment Anything 2 was a breakthrough in image segmentation" ([source](https://news.ycombinator.com/item?id=46061938))
- Accurate zero-shot segmentation with text prompts (3 comments in the sample): "Nowadays that’s no problem anymore as you can just use models like meta’s segment anything that do zero-shot bounding boxes" ([source](https://news.ycombinator.com/item?id=45873919)) "this model is extremely good - the "zero-shot" text prompted detection is a huge step ahead of what we've seen before" ([source](https://news.ycombinator.com/item?id=45985557)) "it does pretty well. At least better than other available models" ([source](https://news.ycombinator.com/item?id=45985159))
- Native streaming video support (2 comments in the sample): "SAM3 has built-in streaming support which is very exciting." ([source](https://news.ycombinator.com/item?id=45985242)) "With the default SAM2 and loading images from the da, we could only process videos with 10^2 - 10^3 frames before running out of memory." ([source](https://news.ycombinator.com/item?id=45986666))
- Teacher model for distilling smaller ones (2 comments in the sample): "SAM3 is probably a great model to distill from when training smaller segmentation models" ([source](https://news.ycombinator.com/item?id=45986456)) "it's so flexible and accurate it's an amazing teacher for lean CNNs" ([source](https://news.ycombinator.com/item?id=45993017))

## What people warn about

- Weak on specialised domains (3 comments in the sample): "It's great with a pony on the beach but really isn't made for more rote industrial type applications." ([source](https://news.ycombinator.com/item?id=45986709)) "These models struggle, understandably so, with medical imaging." ([source](https://news.ycombinator.com/item?id=45991393)) "SAM3 works bad on engineering drawings while SAM2 kinda works" ([source](https://news.ycombinator.com/item?id=46046475))
- Doesn't fit some specific use cases (2 comments in the sample): "It is an amazing model but it doesn't work for my use case" ([source](https://news.ycombinator.com/item?id=45987727)) "This fantastic sam model is maybe not the right fit for our application." ([source](https://news.ycombinator.com/item?id=45989521))

## 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, "Segment Anything Model 3", https://hackerapproved.com/p/meta-segment-anything-model-3/ (38 recommend, 14 mixed, 8 criticize; Hacker News comments, November 2025 to October 2026).

Counts updated 2 Nov 2026, with data through October 31, 2026.