Development & infrastructure

44tools

Supabase

View profile
Development & infrastructureCloud · Self-hosted
Free tierOpen source

What you can do

  • Supabase is an open-source platform built on PostgreSQL that gives an app a database, sign-in, an automatic API, file storage and a web dashboard called Studio. Row-level security policies decide which rows each signed-in person may read or change. It runs as a managed cloud service or self-hosted with Docker on your own server, and app builders such as Lovable are built to work with it.

Before choosing

  • The self-hosted version lacks the cloud's managed backups and point-in-time recovery, branching, advanced metrics and the management API, runs one project per install and is community-supported. The secret key bypasses every row-level security policy and must never reach a browser.

Biopython

View profile
Development & infrastructureDesktop · Self-hosted · Cloud
FreeOpen source

What you can do

  • Biopython is an open-source Python library for computational molecular biology, with modules for sequence handling (Bio.Seq/SeqIO), BLAST search wrappers, protein structure analysis (Bio.PDB), phylogenetics (Bio.Phylo) and NCBI Entrez access. It installs with pip and drops into any Python script, pipeline or notebook.

Before choosing

  • Requires NumPy and a Python 3 environment; it is a scripting library, not a standalone application with a GUI.

Cellpose

View profile
Local & self-hosted appsDevelopment & infrastructureDesktop · Self-hosted · Cloud
FreeOpen source

What you can do

  • Cellpose is an open-source generalist model for segmenting cells and nuclei in microscopy images, distributed as a Python package with a drag-and-drop GUI. The current Cellpose-SAM model runs on CPU or GPU, handles 3D volumes, and supports human-in-the-loop retraining on images the base model gets wrong.

Before choosing

  • GPU is optional but recommended; segmenting many images on CPU alone is slow.

Alibaba Open Code Review

View profile
Development & infrastructureDesktop · Self-hosted
BYOKOpen source

What you can do

  • An open-source code-review CLI that reads Git diffs or whole files and produces structured, line-level review comments using a configurable language model or a host coding agent.

Before choosing

  • Apache-2.0 software; model-provider or host-agent costs are separate. Requires Git 2.41 or newer. The CLI runs on your machine or runner, but model-backed review sends code context to the configured provider. Review findings need human assessment; no accuracy or performance benchmark was independently reproduced.
here.now logo

here.now

View profile
Development & infrastructureCloud
Free tier

What you can do

  • Cloud static hosting for agent-created websites, apps, files, and other assets at live URLs.

Before choosing

  • Free anonymous sites expire after 24 hours; this is static hosting, not a long-running server process.
Ponytail logo

Ponytail

View profile
Development & infrastructureDesktop
FreeOpen source

What you can do

  • An MIT-licensed ruleset for making coding agents write the least code that works.

Before choosing

  • Underlying model or harness costs are separate.

Weaviate

View profile
Development & infrastructureCloud · Self-hosted
Free tier

What you can do

  • Store objects with embeddings for semantic and hybrid search.

Before choosing

  • Plan how embeddings and retrieved records will be produced and maintained.

vLLM

View profile
Development & infrastructureSelf-hosted
FreeOpen source

What you can do

  • Serve language models to many simultaneous users from shared GPU hardware.

Before choosing

  • It is GPU-first and operationally oriented toward concurrent service workloads.

Uptime Kuma

View profile
Development & infrastructureSelf-hosted
FreeOpen source

What you can do

  • Monitor HTTP, TCP, DNS, container, and push-based service checks.

Before choosing

  • Availability checks cannot establish that a workflow's output is correct.

Text Generation Inference

View profile
Development & infrastructureSelf-hosted
Free

What you can do

  • Deploy open language models behind an inference service using Docker.

Before choosing

  • TGI is in maintenance mode, so expect limited new feature development.

Promptfoo

View profile
Development & infrastructureSelf-hosted · Cloud
Free tierOpen source

What you can do

  • Define repeatable evaluations for prompts, models, providers, and datasets.

Before choosing

  • Results are only as useful as the test cases and assertions you define.

Podman

View profile
Development & infrastructureDesktop · Self-hosted
FreeOpen source

What you can do

  • Run OCI containers without a central daemon.

Before choosing

  • podman compose depends on an external provider whose behavior varies by environment.

Pipecat

View profile
Development & infrastructureSelf-hosted
Free

What you can do

  • Assemble voice activity detection, transcription, language, and speech components.

Before choosing

  • You must integrate and self-host the separate pipeline components.

OpenVINO GenAI

View profile
Development & infrastructureSelf-hosted
Free

What you can do

  • Run language, vision, image, speech-to-text, and text-to-speech models locally.

Before choosing

  • Choose compatible models and hardware before building an OpenVINO workflow.

OpenAI Agents SDK

View profile
Development & infrastructureSelf-hosted
Free

What you can do

  • Create text, sandbox, realtime, and voice agents in Python.

Before choosing

  • You need Python 3.10 or newer and must configure a supported model provider.

ONNX Runtime GenAI

View profile
Development & infrastructureSelf-hosted
Free

What you can do

  • Embed compatible ONNX model generation with token-by-token controls in an application.

Before choosing

  • Preview APIs, model artifacts and execution providers must match; it is not a ready-made chat app.
ollama logo

Ollama

View profile
Local & self-hosted appsDevelopment & infrastructureSelf-hosted
FreeOpen source

What you can do

  • Download and run supported models through local chat, CLI and HTTP APIs, with optional cloud features.

Before choosing

  • Memory and GPU fit depend on model and context. Cloud selection changes the data path. Keep the local API on loopback unless you add remote access controls.

NVIDIA System Management Interface

View profile
Development & infrastructureSelf-hosted
Free

What you can do

  • Inspect driver-visible NVIDIA GPU memory, utilization and device state before local inference.

Before choosing

  • NVIDIA-only diagnostics: one sample is not peak-memory evidence and the CLI table is not a stable API.

Neo4j

View profile
Development & infrastructureCloud · Self-hosted
Free tier

What you can do

  • Model connected entities as nodes and relationships.

Before choosing

  • Check edition licensing: Community is GPL v3, while other offerings require commercial terms.

mem0

View profile
Development & infrastructureCloud · Self-hosted
Free tier

What you can do

  • Give assistants retained facts across separate sessions.

Before choosing

  • You must provide an API key and account for model and storage usage.

LocalAI

View profile
Local & self-hosted appsDevelopment & infrastructureSelf-hosted
FreeOpen source

What you can do

  • Self-host compatible APIs and a web interface over selected local AI backends.

Before choosing

  • Every feature needs a compatible backend; remote exposure requires deliberate authentication and network controls.
lmstudio logo

LM Studio

View profile
Local & self-hosted appsDevelopment & infrastructureSelf-hosted
Free

What you can do

  • Download, chat with and locally serve compatible models from a desktop application.

Before choosing

  • Model fit, extensions, updates and selected integrations can create separate resource or network paths.

llm-d

View profile
Development & infrastructureSelf-hosted
Free

What you can do

  • Route Kubernetes model-server pools with inference-aware scheduling and cache-aware routing.

Before choosing

  • Requires a model engine, Kubernetes, accelerator capacity and sustained platform operations.

llm-compressor

View profile
Development & infrastructureSelf-hosted
FreeOpen source

What you can do

  • Compress Hugging Face checkpoints for vLLM deployment.

Before choosing

  • Compression changes model artifacts and requires checking compatibility with your serving setup.