Nebius Customer AtlasNot affiliated · Research by Not affiliated · By @anthonybs
Not affiliated with Nebius · Independent research by @anthonybsSee changes
AllNeedle

Needle

Fine-tuning platform operated by Cactus…

What’s disclosed

Needle's privacy policy lists Nebius among its processors. Nebius supplies GPU and CPU machines that execute runs for the fine-tuning platform.

Commercial use. Needle uses Nebius GPU and CPU machines to execute runs for its fine-tuning platform.

Sources

1 source

Needle fine-tuning platform privacy policy

Listed 29 Oct 2026 · Cactus Compute, Inc. · Web page

Needle's privacy policy lists Nebius as a processor supplying GPU and CPU machines that execute platform runs.

Open original source
Needle fine-tuning platform privacy policyListed 29 Oct 2026 · Cactus Compute, Inc. · Web pageNeedle's privacy policy lists Nebius as a processor supplying GPU and CPU machines that execute platform runs.

Company context

cactuscompute.com

Needle is an AI model and fine-tuning platform operated by Cactus Compute, Inc. It helps developers build software that turns natural-language requests into tool calls and extracts structured information from text. Its Needle 3 models run locally on phones, wearables, computers and embedded devices. Developers can customize models through a Python package or hosted platform, which provides synthetic training data, fine-tuning, evaluation, dataset management and downloadable models for deployment.

Cactus Compute — Needle product page and platform terms

Documented financing · private company

AmountUndisclosedFunding round (stage unspecified)
· Cactus Compute — official documentation roadmap

Recent updates

4 updates
  1. Product launch

    Cactus releases Whistle speech recognition model

    Cactus released Whistle, a 16.9 MB speech recognition model supporting seven languages on CPU-only devices. It shares Needle's inference engine, allowing speech transcription and tool calling to run together locally.

    Cactus Compute
  2. Product launch

    Cactus releases Needle 3 automation model

    Cactus released Needle 3 for local tool calling, structured extraction and text embeddings. Developers can select subnetworks of different depths, with deployment binaries ranging from 8 to 29 MB.

    Cactus Compute
  3. Partnership

    Cactus announces Pebble partnership

    Cactus announced that its technology powers on-device intelligence for Pebble watches and rings.

    Cactus Compute
Next, A–ZNeoDemos
← Explore all customers

Needle

Needle is a fine-tuning platform operated by Cactus Compute, Inc. · Infrastructure

What’s disclosed

Needle's privacy policy lists Nebius among its processors. Nebius supplies GPU and CPU machines that execute runs for the fine-tuning platform.

Commercial use. Needle uses Nebius GPU and CPU machines to execute runs for its fine-tuning platform.

Sources

1 source

Needle fine-tuning platform privacy policy

Listed 29 Oct 2026 · Cactus Compute, Inc. · Web page

Needle's privacy policy lists Nebius as a processor supplying GPU and CPU machines that execute platform runs.

Open original source
Needle fine-tuning platform privacy policyListed 29 Oct 2026 · Cactus Compute, Inc. · Web pageNeedle's privacy policy lists Nebius as a processor supplying GPU and CPU machines that execute platform runs.

Company context

cactuscompute.com

Needle is an AI model and fine-tuning platform operated by Cactus Compute, Inc. It helps developers build software that turns natural-language requests into tool calls and extracts structured information from text. Its Needle 3 models run locally on phones, wearables, computers and embedded devices. Developers can customize models through a Python package or hosted platform, which provides synthetic training data, fine-tuning, evaluation, dataset management and downloadable models for deployment.

Cactus Compute — Needle product page and platform terms

Documented financing · private company

AmountUndisclosedFunding round (stage unspecified)
· Cactus Compute — official documentation roadmap

Recent updates

4 updates
  1. Product launch

    Cactus releases Whistle speech recognition model

    Cactus released Whistle, a 16.9 MB speech recognition model supporting seven languages on CPU-only devices. It shares Needle's inference engine, allowing speech transcription and tool calling to run together locally.

    Cactus Compute
  2. Product launch

    Cactus releases Needle 3 automation model

    Cactus released Needle 3 for local tool calling, structured extraction and text embeddings. Developers can select subnetworks of different depths, with deployment binaries ranging from 8 to 29 MB.

    Cactus Compute
  3. Partnership

    Cactus announces Pebble partnership

    Cactus announced that its technology powers on-device intelligence for Pebble watches and rings.

    Cactus Compute