OctosArmy

OctosArmy x staik
Official Partner

You build autonomous agents with OctosArmy. We give you the GPU muscle. OctosArmy developers get an exclusive Early Agent plan with priority queue on Swedish hardware.

What is OctosArmy?

OctosArmy is an AI-powered file management platform that automates complex tasks on your computer using autonomous agents that collaborate as a hierarchical team.

Autonomous agents

Scout, Brainstormer, Coder, Reviewer — specialized agents that solve tasks together.

Team collaboration

Agents working together, with automatic error correction when a step fails.

Security first

Permission-based access — agents can only reach folders you explicitly allow.

Why staik + OctosArmy?

OctosArmy uses Ollama locally. With staik you can scale up without buying GPU hardware — same OpenAI-compatible API, but with more powerful models on dedicated GPUs.

More powerful models

Run gemma4:31b, qwen3.6:35b and qwen3.5:9b — models on dedicated GPU hardware in Sweden.

Data stays in Sweden

All inference runs on dedicated GPU hardware in Sweden. GDPR-compliant, no data sharing with third parties.

Same API

OpenAI-compatible API — just change the base_url in your OctosArmy configuration. No code changes needed.

Free during partner program

OctosArmy developers get an Early Agent plan completely free. We'll notify you well in advance before terms change.

Exclusive for OctosArmy developers

Early Agent

Free during the partner program

  • 100,000 tokens/hour (rolling window)
  • Priority FIFO queue
  • All models: qwen3.6:35b-a3b, qwen3.5:9b, gemma4:31b
  • OpenAI-compatible API with tool calling
  • Data stays in Sweden (GDPR)
  • 262K token context window

Enter your email to get your API key. Free during the partner program.

Integrate with OctosArmy

TypeScript — OctosArmy agent
import OpenAI from "openai"

const client = new OpenAI({
  baseURL: "https://api.staik.se/v1",
  apiKey: process.env.STAIK_API_KEY
})

// Use as drop-in replacement for local Ollama
const response = await client.chat.completions.create({
  model: "qwen3.6:35b-a3b",
  messages: [{ role: "user", content: "Analyze this file structure" }]
})