If you're looking for a Velvet alternative, Airtrain AI could be worth a look. It's a no-code compute platform geared for data teams that have to wrangle large data pipelines and is built with tools for large language models. It's got an LLM Playground for experimenting with different models, a Dataset Explorer for visualizing and curating data, and AI Scoring for evaluating models. The service can be used to fine-tune and deploy your own AI models, and it's designed to be affordable.
Another contender is Predibase. The service lets developers fine-tune and serve large language models at a low cost. It supports a variety of models and offers free serverless inference for up to 1 million tokens per day. Predibase also offers enterprise-level security and a pay-as-you-go pricing model, so it can be used for a variety of AI tasks.
If you're looking for a broader AI foundation, Abacus.AI offers a suite of tools for building and running AI agents and systems at large scale. It includes tools like ChatLLM for building RAG systems, AI Agents for automating tasks, and predictive modeling tools. Abacus.AI is designed to make AI easier to integrate into applications, with features like high availability, governance and compliance.
Last, Keywords AI offers a unified DevOps platform for building, deploying and monitoring LLM-based AI applications. It has a single API endpoint for multiple models, can handle multiple calls in parallel without a latency penalty, and offers tools for managing prompts and monitoring performance. Keywords AI is geared for AI startups that want to focus on their products without worrying about infrastructure.