13 models, including openai/gpt-oss-120b and meta-llama/llama-3.3-70b-instruct, will be retired at 16:00 UTC on Nov 26, 2026. View the deprecation notice
Paused sandboxes stop CPU and RAM billing. You pay only for persistent storage beyond the free 60 GB per account, at USD 0.00009/GB/h.
Paused sandboxes stop CPU and RAM billing. While paused, a sandbox’s filesystem and memory are kept in persistent storage, which is free up to 60 GB per account and $0.00009/GB/h beyond that.
from novita_sandbox import Novitanovita = Novita()sandbox = novita.sandbox.create()sandbox.pause() # stop CPU and RAM billingsandbox = novita.sandbox.connect(sandbox.sandbox_id) # resume with state intact
import { Novita } from 'novita-sandbox'const novita = new Novita()const sandbox = await novita.sandbox.create()await sandbox.pause() // stop CPU and RAM billingconst resumed = await novita.sandbox.connect(sandbox.sandboxId) // resume with state intact