Qoddi hosts apps from Git or Docker and adds AI inference, custom models, and GPU compute. Free static sites; paid apps from $6/month.
Researched and fact-checked by Bartu Cavusoglu · Updated October 26, 2026
Qoddi is a cloud platform that combines application hosting with AI services. According to its website, teams can deploy websites, APIs, background workers, and databases from Git or Docker, then add managed AI inference, custom model deployments, or NVIDIA GPU compute without leaving the platform. The company groups its offering into App Hosting, AI Inference, Custom Models, and GPU Cloud.
The platform supports Node.js, Python, Go, Ruby, PHP, Java, and Docker-based applications, along with PostgreSQL, MySQL, MongoDB, and Redis. Qoddi also publishes pages positioning itself as a Heroku alternative and a Netlify alternative. Its pricing combines free static sites, per-second app plans with monthly caps, and custom quotes for AI workloads.
Qoddi describes a three-step workflow. First, you connect a Git repository or Docker container and keep the languages and tools you already use. Second, you choose resources, connect a database, and add the services the application needs. Third, you push code and Qoddi handles the infrastructure underneath, including builds, which have no minute limits.
For AI, the workflow is similar. Managed inference runs a model to produce outputs for your application, such as an assistant or smarter search. Custom Models lets you bring your own model and runtime requirements, and the team builds a deployment around them. GPU Cloud provides NVIDIA instances that run Docker workloads for heavier compute needs.
On the hosting side, Qoddi offers Git and Docker deployments, preview environments after each Git push, and paid plans that include daily backups, SSL certificates, load balancing, and autoscaling. App plans also include unlimited bandwidth and unlimited team members. Docker and marketplace apps get 50 GB of persistent storage, and extra storage costs $0.16 per GB per day.
The AI and compute side is broader than most hosting platforms. GPU Cloud offers six NVIDIA configurations, from a quarter A10 with 6 GB of GPU memory to a B200 with 180 GB. Managed inference and custom model deployments are sold separately from GPU infrastructure, so GPU pricing does not include them.
Qoddi is a cloud platform for hosting applications and running AI workloads. It supports deployments from Git or Docker, databases, managed AI inference, custom model hosting, and NVIDIA GPU compute.
Paid app hosting starts at $6/month for the XS plan with 1 GB RAM and 1 vCPU, and goes up to $120/month for XL. GPU plans run from $190 to $5,600 per month. Inference and custom model pricing is quoted per workload.
Yes. The first three static sites are free, and additional static sites cost $3/month per app. Account verification still requires a credit or debit card.
Qoddi supports Node.js, Python, Go, Ruby, PHP, Java, and Docker-based applications. Supported databases include PostgreSQL, MySQL, MongoDB, and Redis.
Inference runs a model to produce outputs for your application. Custom Models is for bringing your own model and shaping its deployment around your workload. The team can help choose a starting point.
Paid app usage is prorated by the second and capped at the published monthly price. The B200 GPU plan is billed monthly. Teams and team members carry no separate seat fee.
Qoddi offers fractional and full NVIDIA A10 and A100 GPUs, an H100, and a B200. GPU memory ranges from 6 GB to 180 GB.
Yes. Preview environments created after a Git push are included at no extra charge. They are deleted after promotion to production or after 48 hours.