# Inception Mercury 2

64 Hacker News commenters recommended Inception Mercury 2 and 28 criticized it, November 2025 to November 2026. It was recommended in 6 of 13 months. It meets the Hacker Approved badge rule.

**Verdict:** Recommended for very fast responses at low cost for agentic subtasks, autocomplete and quick iteration. Critics say it trails frontier models on accuracy and reliability.

## Why people recommend it

- Very fast token speed (8 comments): "The main differentiator is latency — it’s ~5× faster at comparable quality." ([source](https://news.ycombinator.com/item?id=47146445)) "it's still incredibly fast compared to what I'm used to with other models" ([source](https://news.ycombinator.com/item?id=50159524)) "The virtually instantaneous responses of Inception's Mercury models" ([source](https://news.ycombinator.com/item?id=47813533))
- Useful for subagents and subtasks (5 comments): "Even pretty dumb models are useful for running subtasks (especially at this sort of speed)." ([source](https://news.ycombinator.com/item?id=49621669)) "this model and previous are halfway decent and you can do diffusion sub-agents" ([source](https://news.ycombinator.com/item?id=49619091)) "early testing has found the model to be great for coding subagents" ([source](https://news.ycombinator.com/item?id=50156345))
- Cheap per request (3 comments): "It's much less impressive-looking than any of the above, but it cost less than 1/20th of a cent" ([source](https://news.ycombinator.com/item?id=49541558)) "gets roflstomped by Mercury 3 at both cost and performance at similar TPS speed" ([source](https://news.ycombinator.com/item?id=50229280)) "Having a cheap 1,000 tok/s option for coding work is a dream" ([source](https://news.ycombinator.com/item?id=50155574))
- Clear gains over earlier versions (2 comments): "with thinking off, it was considerably better than Mercury 2 and generally good in fact, not very sloppy" ([source](https://news.ycombinator.com/item?id=49623573)) "Seems like a definite improvement over 2.5 though in my testing" ([source](https://news.ycombinator.com/item?id=50155391))

## What people warn about

- Trails frontier models on quality (5 comments): "I can see some promise with diffusion LLMs, but getting them comparable to the frontier is going to require a ton of work" ([source](https://news.ycombinator.com/item?id=47146441)) "Output quality, tool calling reliability and task adherence tend to be far more reliable across all three over Mercury 2" ([source](https://news.ycombinator.com/item?id=47765586)) "it is light years behind frontier models. Speed is great, but wrong is much worse than slow" ([source](https://news.ycombinator.com/item?id=49628666))
- Wrong, incorrect or looping output (3 comments): "it generates incorrect information, and it doesn't fix it when later in the text it realizes that it is incorrect" ([source](https://news.ycombinator.com/item?id=47150118)) "I had a very odd interaction somewhat similar to how weak transformer models get into a loop" ([source](https://news.ycombinator.com/item?id=47146718)) "it is light years behind frontier models. Speed is great, but wrong is much worse than slow" ([source](https://news.ycombinator.com/item?id=49628666))
- Not as fast as advertised (2 comments): "It's not really all that fast, for most things it's full seconds or at least a significant portion of a second." ([source](https://news.ycombinator.com/item?id=50159653)) "It seems like the chat demo is really suffering from the effect of everything going into a queue." ([source](https://news.ycombinator.com/item?id=47145725))
- Expensive without input caching (2 comments): "the lack of an input cache discount makes it prohibitively expensive for most use cases that aren't one-shot prompts" ([source](https://news.ycombinator.com/item?id=49824909)) "Pricing at $0.25 and $0.75 already puts its cost well above reasonably reputable inference providers" ([source](https://news.ycombinator.com/item?id=49823863))

## Alternatives hackers recommend

- [DiffusionGemma](https://hackerapproved.com/p/diffusiongemma.md): 15 recommend
- [Inception Labs](https://hackerapproved.com/p/inception-labs.md): 5 recommend

## Where to get it

- [Visit the official site](https://www.inceptionlabs.ai/)

## Cite as

Hacker Approved, "Inception Mercury 2", https://hackerapproved.com/p/mercury-2/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).