Fine-Tuning & Model Training Platforms
16 companies
0 of 0 categories open
Relationship:
Loading the map…
On this page
Platforms for adapting an existing model to specific data, or training one outright. They run from managed services where a dataset goes in and an endpoint comes out, all the way down to raw distributed-training infrastructure. Membership overlaps heavily with GPU cloud, because several providers sell both the machines and the managed layer above them.
What to compare
Which base models are supported? Do the resulting weights travel with you, or stay locked to the platform? Then compare the fine-tuning methods on offer and how training and serving are priced against each other. Owning the output matters far more than it looks like it will.
16 companies in this category
CerebrasMaker of wafer-scale AI chips used for large-model training.
CoreWeaveSpecialized cloud provider offering large-scale GPU infrastructure for AI.
Crusoe EnergyGPU cloud provider powered by stranded/clean energy sources.
Databricks Mosaic AIDatabricks' unit for large-scale model training and fine-tuning.
Determined AI (HPE)Open-source deep learning training platform, now part of HPE.
FluidStackGPU cloud provider offering large training clusters on demand.
GraphcoreBritish AI chipmaker building IPU accelerators for training and inference.
Lambda LabsGPU cloud and hardware provider focused on AI training workloads.
ModalServerless cloud platform popular for running ML training and inference jobs.
NebiusEuropean AI cloud platform for training and inference infrastructure.
OctoAIModel serving and fine-tuning platform (acquired by NVIDIA).
Paperspace (DigitalOcean)Cloud GPU platform for ML training, now part of DigitalOcean.
PredibasePlatform specialized in efficient fine-tuning and serving of LLMs.
RunPodGPU cloud platform for training, fine-tuning, and serving AI models.
SambaNovaAI hardware and platform company enabling custom model training.
Voltage ParkGPU cloud provider offering large-scale NVIDIA compute clusters.