Why Ollama
Ollama is the easiest way to run open-weight models locally. One binary, model pulls via tags, OpenAI-compatible HTTP endpoint on localhost. Pair Digitorn with Ollama for fully offline coding agents, air-gapped research bots, or compliance-heavy deployments where the data cannot leave your machine. Quality trails frontier models for the hardest tasks but closes monthly.
Models worth knowing about
premium
llama3.3:70b
General-purpose, 70B-class quality
specialty
qwen2.5-coder:32b
Code-tuned, strong on smaller hardware
specialty
deepseek-r1:70b
Reasoning-heavy work, locally
Strengths
- Fully offline, your code and data never leave the box
- Zero cost beyond compute
- OpenAI-compatible endpoint, drop-in
- Model swaps are one CLI command
Worth knowing
- No catalog entry needed (local provider, no credential to vault)
- Quality below frontier on complex tool use
- Latency depends on hardware, M-series Macs and consumer GPUs vary
- Tool-call format is best-effort across model families
Drop into your app.yaml
agent brain block
1brain:2 provider: openai_compat3 model: llama3.3:70b4 config:5 base_url: "http://localhost:11434/v1"6 api_key: "ollama"7 temperature: 0.2⚡ Inline config (catalog support pending)
Ollamadoesn't have a first-class catalog entry yet. Configure it inline using env templates, the same way the blog examples show. Native catalog support is on the roadmap.
Run one in 5 minutes
Install Digitorn and chat with your Ollama agent
# 1. install the runtime
curl -sSL https://digitorn.ai/install | sh
# 2. Ollama runs locally, no credential to register
# 3. install a starter agent and chat
digitorn install hub://digitorn/digitorn-code
digitorn chat digitorn-codeRead the deep dive
Why we chose YAML over Python for our agent runtime
Read article
Newsletter
Get the next post in your inbox.
Engineering notes from the Digitorn team. No marketing, no launch announcements, no "10 prompts that will change your life". Just the things we write that we'd want to read.
Other providers Digitorn supports
frontierAnthropic Claude/integrations/anthropicfrontierOpenAI/integrations/openaifastDeepSeek/integrations/deepseekfrontierMistral/integrations/mistralopenvLLM/integrations/vllmfastGroq/integrations/groqenterpriseAzure OpenAI/integrations/azure-openaifrontierGoogle Gemini/integrations/google-geminiopenTogether AI/integrations/together-ai