Pythia
One local API call gives your agent the entire live state of the planet — every major world event at once, plus forecasts. 40+ free keyless feeds, runs entirely on Ollama. No cloud, no cost. Keyless live feeds into one global view and forecasts what’s likely next. No cloud, no keys, no cost. 1d,1w,1m,1y predictions by Mirofish.
What is it?
What it is
PYTHIA is a local, open-source oracle that fuses the MiroFish swarm-intelligence prediction engine and the Osiris live global-intelligence globe into a single system that ingests real-time worldwide data (news, conflict, weather, markets, etc.) and generates forecasts for 24h, week, month, and year horizons, complete with reasoning, probability, and location, while providing a machine-readable API for agents.
Why it exists
The world constantly broadcasts its future through countless signals, but no single observer can monitor and reason across all of them. PYTHIA exists to give users and AI agents a unified, keyless, on‑premise view of the planet’s live state and its likely evolution, enabling informed decisions, real‑world context, and automated reactions to global changes.
Who should use it
Developers building AI agents, data scientists, researchers, and analysts who need real-time global event data and forecasts for decision-making or agent context.
Who should avoid it
Non-technical users seeking a ready-to-use forecasting SaaS, those unwilling to install and run local dependencies (Ollama, Osiris, Python), or anyone looking for a point-and-click interface without coding.
How it works
A quick walkthrough in plain English
How Pythia works
Step 1 of 3
You interact with it
Open Pythia, send a request, or connect it to your stack.
Features
Advantages
- Runs entirely on local hardware – zero cloud cost or latency
- No API keys or external accounts – privacy and security
- Immediate, up‑to‑date forecasts from live feeds
- Swarm deliberation provides both consensus and dissent, improving reliability
- Built‑in performance metrics allow continuous improvement
- What‑if analysis enables scenario planning
- Webhook integration supports real‑time event handling
- Open API and MCP support easy integration with agents
- Model flexibility via Ollama allows tailoring to resources
- Extensible architecture with separate modules (engine, osiris overlay, integrations)
- Free, open‑source components (MIT license)
Disadvantages
- Requires significant local compute and memory for LLM and swarm
- Complex setup: needs Ollama, Osiris checkout, Python, uv, etc.
- Dependent on local hardware; scaling limited compared to cloud
- Accuracy depends on quality of feeds and LLM reasoning
- No built‑in persistence beyond local ledger – data loss on crash if not backed up
- Limited to 24 h–1 yr horizons; longer forecasts not supported
- Requires manual configuration of feeds and models
- Potential performance bottlenecks in real‑time sensing loop
Installation
native
Prerequisites: install Ollama and pull a chat model (e.g., `ollama pull llama3.1`), clone an Osiris checkout and apply the PYTHIA overlay (see `integrations/osiris/INSTALL.md`), install Python 3.11+ and uv. Then run: 1. `git clone https://github.com/jangles-byte/Pythia && cd Pythia` 2. `cp .env.example .env` (configures defaults; no API keys needed) 3. `./run-all.sh` (starts the Osiris globe on :3000 and the engine API on :8088, opening the UI). On macOS you can also double‑click `PYTHIA.app` to launch the same stack.
FAQ
What is Pythia?
Pythia is an open-source project licensed under MIT.
What language is Pythia built with?
Primary language: TypeScript.
Is it free to use?
Yes. Licensed under MIT. Check the license for commercial use.
Featured in Videos
YouTube tutorials and walkthroughs for Pythia
t5
70,223,459 views
Starts at 00:00
Alternatives
Similar projects ranked by category, topics, and text overlap.
