Luna 1.2.3

Warning

🚨 Breaking Changes Included – Released on July 21, 2026

⚠️ Breaking Changes

Use-Case Problem Library Moved to luna-usecases The bundled problem library under luna_quantum.solve.use_cases has been removed and now lives in the standalone luna-usecases package on PyPI. It provides 45+ ready-to-use optimization problems — TSP, MaxCut, Knapsack, Portfolio Optimization, Job Shop Scheduling, and many more — with helpers to generate instances, formulate a Model, and interpret the result.

Bash

pip install luna-usecases

Python

from luna_usecases.integer_knapsack_problem import IkpData, IkpFormulation, IkpInstance
from luna_quantum import algorithms

data = IkpData.generate_random(n_items=20, capacity=50, seed=42)
instance = IkpInstance(data=data, formulation=IkpFormulation())

model = instance.formulate()
result = algorithms.SCIP().run(model).result()
solution = instance.interpret(result)
print(solution.print())

🛠️ Improvements

Faster Solve-Job Polling Waiting for a solve job to finish is now noticeably quicker. The first status check happens almost immediately, and the wait between checks grows with capped exponential backoff instead of a fixed increment — so short jobs return sooner while long-running jobs still back off gracefully. The progress output has also been reworked to render cleanly in Jupyter notebooks.
Python
```
solution = solve_job.result(
    sleep_time_initial=0.5,  # first check after 0.5s
    backoff_factor=2.0,      # double the wait after each check
    sleep_time_max=30.0,     # capped at 30s between checks
)
```
More Resilient Connections The client now automatically retries transient connection errors when talking to the Luna platform, making solves more robust against brief network hiccups.