Engineering Blog | Cactus cactuscompute.com

Deep dives into on-device AI, inference optimization, and running models on smartphones, laptops, and edge hardware.

Engineering Blog·

Needle: We Distilled Gemini Tool Calling into a 26M Model

Covers 2 stories See all stories this covers including cactus-compute/needle: 26m function call model that runs on incredibly small devices
Feeds
Engineering Blog·

TurboQuant-H: Hadamard Rotation for 2-Bit Embedding Quantization

Covers 2 stories See all stories this covers including TurboQuant: Online Vector Quantization with Near-optimal Distortion Rate
Feeds
Engineering Blog·

LFM-2.5-350m on Cactus: 140 tok/sec, Single Core, 355 MB

Feeds
Engineering Blog·

Sub-150ms Transcription with Cloud-Level Accuracy: Why We Built a Hybrid Engine

Covers Framework for on-device inference on mobile phones.
Feeds
Engineering Blog·

Ridiculously Fast On-Device Transcription: Reviewing Parakeet CTC 1.1B with Cactus

Feeds
Engineering Blog·

The Sweet Spot for Mac Code Use: Reviewing LFM2 24B MoE A2B with Cactus

Feeds

Keyboard Shortcuts

Navigation

Next / previous post
j/k
Open post
oorEnter
Preview post
v

Post Actions

Love post
a
Like post
l
Dislike post
d
Undo reaction
u
Save / unsave
s

Recommendations

Add interest / feed
Enter
Not interested
x

Go to

Home
gh
Interests
gi
Feeds
gf
Likes
gl
History
gy
Changelog
gc
Settings
gs
Discover
gb
Search
/

General

Show this help
?
Submit feedback
!
Close modal / unfocus
Esc

Press ? anytime to show this help