MyAI GPU Agent
Earn MYAI tokens by sharing your GPU compute power with the MyAI distributed inference network.
Your GPU runs local AI inference via Ollama. Jobs are dispatched from the MyAI coordinator and you earn MYAI on Base for every completed job.
Quick Install
macOS / Linux:
curl -fsSL https://myaitoken.io/install | bash
With your wallet address (to start earning immediately):
WALLET=0xYourWalletAddress curl -fsSL https://myaitoken.io/install | bash
pip (any platform):
pip install myai-agent
myai-agent install --wallet 0xYourWalletAddress
Requirements
| Requirement | Version | Notes |
|---|---|---|
| Python | 3.8+ | Zero external deps — pure stdlib |
| Ollama | Any | Local inference engine |
| GPU | Optional | CPU works too, just slower |
Supported GPUs: NVIDIA (CUDA), Apple Silicon (Metal), AMD (ROCm)
Manual Setup
1. Install Ollama
# macOS
brew install --cask ollama
# Linux
curl -fsSL https://ollama.com/install.sh | sh
# Windows
# Download from https://ollama.com/download
2. Pull a model
ollama pull llama3.2 # recommended (2GB, fast)
ollama pull llama3.1:8b # larger, better quality
ollama pull mistral # alternative
3. Install the agent
pip install myai-agent
4. Run or install as service
One-off run:
myai-agent start --wallet 0xYourWallet
Install as service (auto-starts on login):
myai-agent install \
--wallet 0xYourWallet \
--name "My GPU Rig" \
--model llama3.2
Commands
myai-agent start Run the agent (foreground)
myai-agent install Install as OS service
myai-agent uninstall Remove OS service
myai-agent status Check service status + agent ID
myai-agent logs Tail agent logs
myai-agent models List available Ollama models
myai-agent run-job <prompt> Test local inference
myai-agent --version Print version
Options (start / install)
| Flag | Default | Description |
|---|---|---|
--coordinator URL |
https://api.myaitoken.io |
Coordinator endpoint |
--ollama URL |
http://localhost:11434 |
Ollama endpoint |
--wallet 0x... |
— | Base EVM wallet for MYAI rewards |
--name NAME |
hostname | Display name for this agent |
--model MODEL |
llama3.2 |
Default Ollama model |
--required-models M1,M2 |
bonsai-8b:latest |
Comma-separated models to auto-pull on startup |
Auto Model Pull
The agent automatically pulls required models on startup if they aren't already present in Ollama.
Default: bonsai-8b:latest is pulled on every agent start if missing (~1.1 GB, 1-bit quantized).
Customize via env var:
REQUIRED_MODELS=bonsai-8b:latest,deepseek-r1:7b myai-agent start
Disable auto-pull:
REQUIRED_MODELS="" myai-agent start
The pull streams progress to the log and blocks registration until complete — so the coordinator always sees accurate model lists.
How It Works
- Agent registers with the coordinator and reports available GPUs + models
- Coordinator receives inference jobs from users
- Agent polls for pending jobs (
GET /api/v1/agents/{id}/jobs/pending) - Agent runs inference locally via Ollama (
POST /api/generate) - Results reported back to coordinator (
POST /api/v1/agents/{id}/jobs/{job_id}/complete) - MYAI tokens distributed to your wallet on Base (Phase 3)
Parallel inference: Large jobs are split across multiple agents simultaneously for faster results.
Platform Notes
macOS
- Installed as a launchd service (
~/Library/LaunchAgents/io.myaitoken.agent.plist) - Logs:
~/Library/Logs/myai-agent/agent.log - Ollama recommended:
brew install --cask ollama
Linux
- Installed as a systemd user service (
~/.config/systemd/user/myai-agent.service) - Logs:
journalctl --user -u myai-agent -f
Windows
- Installed as a Task Scheduler task (runs on login)
- Download Ollama from ollama.com/download
Configuration
Environment variables (set in OS service config, or export before myai-agent start):
| Variable | Default | Description |
|---|---|---|
COORDINATOR_URL |
https://api.myaitoken.io |
Coordinator API URL |
OLLAMA_URL |
http://localhost:11434 |
Ollama server URL |
AGENT_NAME |
hostname | Display name |
AGENT_WALLET |
— | Base wallet for MYAI rewards |
POLL_INTERVAL |
5 |
Seconds between job polls |
HEARTBEAT_INTERVAL |
30 |
Seconds between heartbeats |
MYAI Token
- Contract:
0xaff22cc20434ce43b3ea10efe10e9360390d327c - Chain: Base (Chain ID: 8453)
- Symbol: MYAI | Supply: 1,000,000,000
License
MIT © MyAI Token
Metadata
Release files for myai-agent 2.0.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| myai_agent-2.0.0.tar.gz | 16.7 kB | Without provenance | Details for myai_agent-2.0.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| myai_agent-2.0.0-py3-none-any.whl | Python 3 | none | any | Details for myai_agent-2.0.0-py3-none-any.whl |
Total release size: 33.1 kB
Release files / myai_agent-2.0.0.tar.gz
| Download URL | myai_agent-2.0.0.tar.gz |
|---|---|
| Size | 16.7 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
c67bc6a1ff4a61140aa529322a0fe9a900561397dd84926e055157df633633b8
|
|
BLAKE2b-256 checksum How to use checksums |
2b45df69d8a749eec9c3e6abfea9582ceb044c3fac5716208b8c2860995f3114
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.10.12
|
Release files / myai_agent-2.0.0-py3-none-any.whl
| Download URL | myai_agent-2.0.0-py3-none-any.whl |
|---|---|
| Size | 16.4 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
d54a83e538e83b64939d99260fae5740bd555bae98baaf27848cd6f0e3a2bfa7
|
|
BLAKE2b-256 checksum How to use checksums |
0ac2e12f0f7fbe883a64c112eba3ac18e2e0a078ad9ec04a86891d4bb93d3aa1
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.10.12
|
Security for myai-agent 2.0.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.