# Production level deployment
**URL:** https://forum.algorand.co/t/production-level-deployment/7348
**Category:** General
**Created:** [July 6, 2022, 5:53am UTC](https://forum.algorand.co/t/production-level-deployment/7348 "2022-07-06T05:53:38Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@isha](https://forum.algorand.co/u/isha)
#### Post date: [July 6, 2022, 5:53am UTC](https://forum.algorand.co/t/production-level-deployment/7348/1 "2022-07-06T05:53:38Z")
Hi all,
I am working on a projet using sandbox environment.
Can anyone guide me with the process I will have to do to take this project to production level on mainnet
Is there any link I can refer?
---
### Author:  [@evertdiazb](https://forum.algorand.co/u/evertdiazb)
#### Post date: [July 11, 2022, 8:52pm UTC](https://forum.algorand.co/t/production-level-deployment/7348/2 "2022-07-11T20:52:43Z")
If you’ve already built your complete dapp on local sandbox, you have 2 options to interact and deploy on mainnet depending if you’re using CLI or the SDK:
- Use the sandbox to connect mainnet: `./sandbox up mainnet`
- Use a third party service API like purestake or similar, just change the algod URL and token. Here you can find a list with available third party nodes: [LINK](https://developer.algorand.org/ecosystem-projects/?tags=api-services)
---
### Author:  [@fabrice](https://forum.algorand.co/u/fabrice)
#### Post date: [July 11, 2022, 9:08pm UTC](https://forum.algorand.co/t/production-level-deployment/7348/3 "2022-07-11T21:08:57Z")
It depends what you mean by _production_ and the type of project.
**If you’re creating something like an NFT** which essentially just requires you to make transactions, then @evertdiazb is the simplest solution.
(Note that even then, sandbox uses fast catchup in a centralized way (the catchup label is provided by Algorand Inc): for full decentralization, you want to sync from scratch or use a catchup label verified by multiple parties. That being said, API services are even more centralized from this point of view. This is something that is true for all blockchains.)
However, **if you are running a full dApp that users can access and use directly** (like [https://tinyman.org/](https://tinyman.org/), [https://www.algofi.org/](https://www.algofi.org/), …), then I would strongly recommend **NOT using sandbox** in production.
sandbox is really just for development or for sending your own transactions (with the caveat specified above).
To move to production in such a setting, you need you need a reliable node.
1. If you have expertise and capacity to run reliable infrastructure, you can run your own node. You may also use providers (see bottom of [https://algorand.foundation/algorand-protocol/network](https://algorand.foundation/algorand-protocol/network)).
Note that if you require very high SLA and your dApp is small, this solution is most likely too costly.
2. If you do not have the expertise to run your own nodes or if it is not worth the cost, the best solution is to use an API service: [Ecosystem Tools & Projects | Algorand Developer Portal](https://developer.algorand.org/ecosystem-projects/?tags=api-services). While API services have free tiers, we highly recommend you to check whether the SLA and customer service provided by the free tier match your need. For high-volume high-value projects, it is most likely necessary to pay for the API service. (This is the same on all blockchain: very high SLA and 24/7 customer services is labor intensive and is costly to achieve. See also [How big can the Algorand Blockchain get before problems start occurring? - #6 by fabrice](https://forum.algorand.co/t/how-big-can-the-algorand-blockchain-get-before-problems-start-occurring/7167/6))
---
### Author:  [@isha](https://forum.algorand.co/u/isha)
#### Post date: [July 12, 2022, 11:00am UTC](https://forum.algorand.co/t/production-level-deployment/7348/4 "2022-07-12T11:00:43Z")
I want to deploy a DAO application on the mainnet.
I’ll explore the options you both have give thanks a lots @evertdiazb and @fabrice!