A lightweight AI application built with Vuetify, Lightning AI and Meteron. Showcases queueing, asset loading and generation.
Rapidly build AI products
Meteron handles load-balancing, ordering, storage and limits for your AI systems.
All-in-one AI toolset
Free developers from time-consuming, unnecessary processes that slow your work, so you and your team can focus on creating
Kickstart your project
An ecosystem of integrations
Start for free, upgrade at any time.
Check out examples on how to build your own projects with Meteron.
An advanced, multi-tenant example where users can register and generate their assets. Assets, queueing and metering are enforced by Meteron. Generation is done by replicate.com.
A collection of individual function on how to send image generation requests through Meteron, query results, ensure per-user limits and more.
Do I need to use any special libraries when integrating Meteron?
How do I tell Meteron where my servers are?
You can do it through the web UI if your servers are static or rarely change. However, we also provide a simple API that you can use to update your servers dynamically. This API can be used to update your servers in real time, for example, when you are using a AI platforms such as lightning.ai, runpod.io, etc.
How does the queue prioritization work?
By default Meteron provides several standard business rules. With each request you can specify priority class (high, medium, low) where high are your VIP users and will not incur any queueing delays. Medium priority class users will incur delays but will always be ahead of low priority users. Low priority requests will be served last, these can be your "free" users that are served when there is no load on the system.
Do I need coding knowledge to use this product?
Meteron is a "low-code" service where some knowledge about HTTP is needed. However, we do provide example on how to integrate Meteron. If you do get stuck, join our Discord server and we will be happy to help out.
Can I host Meteron server myself?
Yes, on-prem licenses are available. You will get a batteries included system that you can run on any cloud provider. Contact us for more info at [email protected].
What forms of payment do you accept?
We accept all major credit cards as well as direct wire transfers.
How does per-user metering work?
When adding model endpoints in Meteron (clusters) you can specify daily and monthly limits. WHen these limits are specified, each time you send an image generation request, add the X-User header with the user ID (or email) and we will ensure that this user cannot go above limits.