Roman Shemet

Mobile AI @ Cactus (YC S25)

[Roman Shemet @RomanShemet]

Plug in local, free, private AI into your mobile apps 🌵

Cactus now supports @FlutterDev, @reactnative , and @kotlin bindings.

+ with function calling, you can also deploy complex agentic pipelines directly in-app.

Fully open-source 🌵 link in bio 🔗

[Aristotelis Economides @aristotelis_eco]

https://t.co/8JsKtSDnqM

[Roman Shemet @RomanShemet]

All cloud fallback is 𝗙𝗥𝗘𝗘 𝘁𝗵𝗶𝘀 𝗙𝗲𝗯𝗿𝘂𝗮𝗿𝘆. Seriously. Make us regret this 😅

We just launched Hybrid Cloud inference and we're too excited for you to try it.

1. Go to https://t.co/341YSbH19V
2. Sign up and create a key
3. Run unlimited on-device transcription and LLM inference with cloud fallback

Cactus Hybrid Cloud runs inference on-device by default, as always. If the on-device model struggles, it automatically hands off inference to the cloud.

Demo for yourself:

𝗯𝗿𝗲𝘄 𝗶𝗻𝘀𝘁𝗮𝗹𝗹 𝗰𝗮𝗰𝘁𝘂𝘀-𝗰𝗼𝗺𝗽𝘂𝘁𝗲/𝗰𝗮𝗰𝘁𝘂𝘀/𝗰𝗮𝗰𝘁𝘂𝘀
𝗰𝗮𝗰𝘁𝘂𝘀 𝘁𝗿𝗮𝗻𝘀𝗰𝗿𝗶𝗯𝗲

[Roman Shemet @RomanShemet]

@dshukertjr Are there best practice guidelines for syncing local<>prod migrations with supabase?

[Roman Shemet @RomanShemet]

@ayusrjn @yoheinakajima @RunAnywhereAI @cactuscompute thank you for the feedback! Could you share which phone & version of the app you used? We'll patch this up for you

[Roman Shemet @RomanShemet]

@btconometrics @Raspberry_Pi you can run @cactuscompute on any raspberry pi.

Cactus also runs zero-copy memory mapping, so you're not constrained by the 8GB RAM

[Roman Shemet @RomanShemet]

@supernalmystic @Raspberry_Pi you can run @cactuscompute on any raspberry pi.

Cactus also runs zero-copy memory mapping, so you're not constrained by the 8GB RAM

[Roman Shemet @RomanShemet]

Which personalization experiences would you want to see in local/on-device AI?

[Roman Shemet @RomanShemet]

@kodjima33 @omidotme on-device AI? Something like @cactuscompute

[Jakub Mroz @jakmroo]

At Cactus🌵, we want on-device AI inference to be as fast as possible - that’s why we decided to use Nitro for our React Native SDK. The performance is insane⚡️and the DX is even better. Making heavy use of the object-oriented Hybrid Objects, and mixing C++ with Swift and Kotlin feels like a breeze.

Thanks @mrousavy - looking forward to shipping more features with Nitro 🚀

[Roman Shemet @RomanShemet]

@SzymonRybczak 😲 who undercut us?

[Marc @mrousavy]

Check out Cactus AI!

[Sélim @SelimBenayat]

Hackathon alert! London, SF, Boston. This Friday! 👀

@nothing is teaming up with @cactuscompute and @huggingface to hack on redefining on-device AI experiences!

Come build something memorable, meet the teams, and ship in 24 hours!

Signups are wild so far 🔥

[Cactus @cactuscompute]

Cactus React Native v1 is live!

Deploy AI on-device with text inference, tool calling, embeddings and more – powered by the fastest edge inference engine 🌵

Our React Native bindings run on @margelo_com's Nitro Modules, yielding the fastest mobile inference we've seen so far.

[Roman Shemet @RomanShemet]

why does every SF startup show off their round on a NYC billboard?

@SF, what’s our Times Square?

[Henry Ndubuaku @Henry_Ndubuaku]

More benchmarks for LFM2 models by @liquidai on Cactus (YC S25).

- When we get to INT4, file sizes should reduce 2x, speed increase 2x, and battery drain reduce 2x. For lossless quantisation, the model should always be post-trained for QAT in a specific way, a trick we mastered in my last role.

- NPUs are 5-11x energy-efficient and up to 10x faster for long-context, after we merge those, you should be able to design complex multi-agent workflows with large contexts safely on phones.

Many ther tricks up our sleeves. Sit back, relax and watch!

[Roman Shemet @RomanShemet]

@v0 is the closest we've come to AGI. Alarming levels of self-awareness 🚨

[Roman Shemet @RomanShemet]

@alexinexxx hope it’s running on @cactuscompute

[Roman Shemet @RomanShemet]

@tomgreenwald @cpaik ✋🏼 we are indeed the guys! –

https://t.co/lKQgufSd1Y

how can we help

[Roman Shemet @RomanShemet]

@janaksunil @ycombinator @usebearai @siddhantpa1iwal congratulations!

👋 Fellow S25 founder here, would love to learn more about your product