Managing access to multiple AI providers across a team or product usually means juggling separate API keys, separate billing, and separate rate limits for each one. New API solves that by sitting in front of all of them as a single, self-hosted gateway: your application talks to one OpenAI/Claude/Gemini-compatible endpoint, and New API handles routing, tracking and billing behind the scenes. New API is an open-source (AGPLv3) self-hosted hub that converts requests between OpenAI, Claude and Gemini API formats, supports 10+ underlying LLM providers, and adds organization-level authentication, permission management, and per-user usage analytics and cost accounting. It includes intelligent routing with automatic retry on failure, Docker-based deployment, a multi-language admin UI, and built-in billing integration (Stripe, EPay) for teams that need to charge internal or external users for AI access. It's a very actively developed project with over 46,000 GitHub stars.
| # | Tool | Best for | Price |
|---|---|---|---|
| 1 | Développeurs | — | |
| 2 | Développeurs et équipes voulant un assistant de code IA auto-hébergé avec mémoire persistante | — | |
| 3 | Équipes IT et sécurité d'entreprises déployant des agents IA internes construits par leurs employés | — | |
| 4 | Développeurs construisant des agents IA en production nécessitant un contrôle fin de l'état et de la fiabilité | — | |
| 5 | Équipes produit et développeurs construisant des chatbots, agents IA ou pipelines RAG sans vouloir tout coder à la main | — | |
| 6 | Développeurs et équipes construisant des systèmes IA ancrés dans leurs propres documents (RAG), de la startup à la grande entreprise | — | |
| 7 | Développeurs et équipes techniques construisant des agents IA autonomes longue durée (recherche, code, tâches multi-étapes) | — | |
| 8 | Développeurs et équipes qui font tourner des agents de code IA (Claude Code, Cursor, Codex...) de façon semi-autonome sur une vraie codebase | — | |
| 9 | Organisations en secteurs réglementés (santé, droit, finance, gouvernement) qui ne peuvent pas envoyer leurs données à un cloud tiers | — | |
| 10 | Développeurs et ingénieurs DevOps qui veulent des agents IA légers et composables intégrés à des pipelines existants (cron, git hooks, scripts) | — | |
| 11 | Chercheurs, travailleurs du savoir et développeurs qui veulent visualiser et contrôler précisément le contexte envoyé à un LLM sur des conversations longues | — | |
| 12 | Créateurs de contenu, entrepreneurs et petites équipes qui vivent dans Telegram et veulent un assistant IA avec mémoire long terme et souveraineté des données | — |
A self-hosted AI coding assistant that remembers past sessions and can act across your terminal, IDE and other tools with a large context window.
Lets IT and security teams see, approve and monitor every internal AI agent employees build, instead of agents spreading unchecked across the company.
A code library from the LangChain team for building AI agents that can pause, remember where they were, and pick back up later — even after a server restart — instead of losing all progress if something goes wrong mid-task.
An open-source platform for building AI chatbots and automated workflows by connecting blocks on a canvas, instead of writing code from scratch.
An open-source toolkit for building AI systems that answer questions using your own documents — feeding an AI model your company's PDFs, wikis, or support tickets so it answers from real content instead of guessing.
An open-source framework from ByteDance for building AI agents that can spend minutes or hours working through a task — researching, writing code, and spinning up helper agents — rather than answering in one quick reply.
A guardrail system that limits which tools an AI coding agent can use at each stage of a task — read-only while planning, edit access only while implementing — so it can't take an action outside its current job.
A free, open-source, self-hosted AI agent platform built for organizations that can't send data to a third-party cloud — think law firms, clinics or banks — with role-based access, memory, and support for every major AI model.
A free, open-source command-line tool that lets you define small, single-purpose AI agents as simple config files and run them like Unix programs — triggered from the terminal, a cron job, a git hook, or piped straight into another command.
A free, open-source infinite canvas that turns your conversations with AI models into an editable graph — so you can see, prune, and reuse exactly which parts of the conversation the AI is actually reading, instead of a single ever-growing chat thread.
A self-hosted AI assistant that lives inside Telegram on your own server, with memory, emotional awareness, and dozens of built-in skills — useful if you want an AI that remembers your life and work without handing your data to a big tech company.
