AI Models
663 tools
663 tools · page 24 of 28
Fireworks AI hosts optimized, low-latency inference for open-source language and multimodal models, aimed at developers building production apps that need speed at scale.
FlowGPT is a community where users discover, share, and rate prompts for language models across many use cases, aimed at people who want proven prompts rather than writing their own.
Forefront provides a chat interface with access to several different underlying AI models, aimed at users who want to compare or switch between models in one place.
GPTLocalhost is a Microsoft Word add-in that connects to language models running locally on a user's own machine, aimed at writers who want AI assistance without sending text to the cloud.
Gretel generates synthetic datasets that preserve the statistical patterns of real data without exposing sensitive information, aimed at teams training models where real data is scarce or private.
Groq built custom chips (LPUs) designed specifically for fast language model inference, and offers a cloud API that developers use when response speed is the top priority.
H2O.ai provides automated machine learning tools that let teams build predictive models with less manual tuning, alongside newer generative AI and LLM tooling.
Haystack, built by deepset, is an open-source Python framework for building search, question-answering, and retrieval-augmented generation pipelines on top of language models.
HiddenLayer monitors deployed machine learning models for adversarial attacks and misuse in real time, aimed at enterprises protecting AI systems in production.
All-in-one AI platform with 500+ AI tools and top models under one subscription
Jan is an open-source desktop app for chatting with local, offline AI models, aimed at users who want a private ChatGPT-style interface without sending data to the cloud.
Kazimir.ai indexes and lets users search across a large collection of AI-generated images, aimed at creators looking for style inspiration or examples of what different models produce.
Krutrim, founded by the team behind ride-hailing company Ola, builds AI language models and cloud infrastructure aimed at serving India's languages and market needs.
LAIKA trains a personalized AI model on a writer's own past work, then generates suggestions and continuations meant to match that writer's individual style rather than a generic voice.
Lakera provides real-time threat detection for AI applications, guarding against prompt injection and other attacks aimed at manipulating language model behavior.
Lambda provides cloud GPU instances and on-prem hardware specifically configured for deep learning workloads, aimed at AI labs and researchers training large models.
LangChain provides building blocks for chaining together language model calls, tools, and data sources into full applications, widely used as a starting point for AI agent projects.
LangSmith traces, evaluates, and monitors language model application runs, aimed at developers who need visibility into why an LLM app produced a particular output in production.
Le Chat is Mistral AI's consumer-facing chatbot, offering conversation, coding help, and image generation built on Mistral's own language models, positioned as a European alternative to US chat assistants.
LightOn builds generative AI models and infrastructure for enterprises that need to run AI privately on their own data, positioned as a French and European alternative to US AI providers.
LiteLLM lets developers call over a hundred different LLM APIs using a single, OpenAI-compatible format, simplifying the process of switching or comparing model providers.
LlamaIndex helps developers ingest, index, and query private data sources so language models can answer questions grounded in that data, a common building block for RAG applications.
LM Studio lets users download and run open-source language models directly on their own computer through a simple desktop interface, without needing cloud API access.
LMQL is a programming language that lets developers write structured, constrained queries against large language models, aimed at more reliable and controllable LLM outputs than plain prompting.