Skip to main content

Python SDK for HyperCLI - GPU orchestration and HyperAgent API

Project description

HyperCLI SDK

Python SDK for HyperCLI - GPU orchestration API.

Installation

pip install hypercli-sdk

Setup

Set your API key:

export HYPER_API_KEY=your_api_key

Or create ~/.hypercli/config:

HYPER_API_KEY=your_api_key

Or pass directly:

client = HyperCLI(api_key="your_api_key")

Usage

from hypercli import HyperCLI

client = HyperCLI()

# Check balance
balance = client.billing.balance()
print(f"Balance: ${balance.total:.2f}")
print(f"Rewards: ${balance.rewards:.2f}")

# List transactions
for tx in client.billing.transactions(limit=10):
    print(f"{tx.transaction_type}: ${tx.amount_usd:.4f}")

# Create a job
job = client.jobs.create(
    image="nvidia/cuda:12.0",
    command="python train.py",
    gpu_type="l40s",
    gpu_count=1,
)
print(f"Job ID: {job.job_id}")
print(f"State: {job.state}")

# List jobs
for job in client.jobs.list():
    print(f"{job.job_id}: {job.state}")

# Get job details
job = client.jobs.get("job_id")

# Get job logs
logs = client.jobs.logs("job_id")

# Get GPU metrics
metrics = client.jobs.metrics("job_id")
for gpu in metrics.gpus:
    print(f"GPU {gpu.index}: {gpu.utilization}% util, {gpu.temperature}°C")

# Cancel a job
client.jobs.cancel("job_id")

# Extend runtime
client.jobs.extend("job_id", runtime=7200)

# Get user info
user = client.user.get()
print(f"User: {user.email}")

HyperAgent API

Use client.agent for discovery and plan metadata, and point the OpenAI SDK at the HyperClaw inference base URL for chat completions:

from hypercli import HyperCLI
from openai import OpenAI

sdk = HyperCLI(api_key="hyper_api_key", agent_api_key="sk-agent")
plans = sdk.agent.plans()

client = OpenAI(
    api_key="your_hyperagent_api_key",
    base_url="https://api.hypercli.com/v1"
)

response = client.chat.completions.create(
    model="deepseek-v3.1",
    messages=[{"role": "user", "content": "Hello!"}]
)

OpenClaw Agents

OpenClaw uses the generic deployment launch surface. registry_url, registry_auth, sync_root, and sync_enabled are generic deployment options; the OpenClaw helpers only add defaults such as routes, image, and sync_root=/home/node.

agent = client.deployments.create_openclaw(
    name="docs-demo",
    start=True,
    heartbeat={"every": "0m"},  # disable upstream OpenClaw heartbeat runs
    registry_url="git.nedos.co",
    registry_auth={"username": "ci", "password": "token"},
)

heartbeat maps directly to upstream OpenClaw config at config.agents.defaults.heartbeat. Omit it to keep upstream defaults, or pass values such as heartbeat={"every": "1h", "target": "last"}.

Error Handling

from hypercli import HyperCLI, APIError

client = HyperCLI()

try:
    job = client.jobs.get("invalid_id")
except APIError as e:
    print(f"Error {e.status_code}: {e.detail}")

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

hypercli_sdk-2026.4.18.post1.tar.gz (102.6 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

hypercli_sdk-2026.4.18.post1-py3-none-any.whl (83.2 kB view details)

Uploaded Python 3

File details

Details for the file hypercli_sdk-2026.4.18.post1.tar.gz.

File metadata

  • Download URL: hypercli_sdk-2026.4.18.post1.tar.gz
  • Upload date:
  • Size: 102.6 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.13

File hashes

Hashes for hypercli_sdk-2026.4.18.post1.tar.gz
Algorithm Hash digest
SHA256 58042f08703e0c3efc33f47ac4ed64f974bddc972da892e07e7d8fc63cfccebf
MD5 a00091b31011e4d6a6620b823a51c140
BLAKE2b-256 83ea747f970db2b4a4d8188019335dc52f894806612fef0a466df255440c76c4

See more details on using hashes here.

File details

Details for the file hypercli_sdk-2026.4.18.post1-py3-none-any.whl.

File metadata

File hashes

Hashes for hypercli_sdk-2026.4.18.post1-py3-none-any.whl
Algorithm Hash digest
SHA256 516502bb8e197779ea9f4a4439cc2b43a2f3b4d766edf76d84ff6263fe24e88b
MD5 15e7d0169e3f26a6e8ef12d23afbd7b6
BLAKE2b-256 662bd59bf1fdb46ccc95f20f10f795539c264e0186047011f68a6023efb4cd1e

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page