LIVE
Mistral Makes Large 4 Available as Open Weights29/10/26 · Mistral AI|Gemini 4 Carbon tops the Artificial Analysis Intelligence Index v529/10/26 · Google|Mistral Large 4 now available on Hugging Face29/10/26 · Mistral AI|Google Introduces "Gemini 4 Carbon" as Next Step in Its Model Lineup28/10/26 · Google|DeepMind unveils Genie 4, a new generation of world model27/10/26 · Google DeepMind|Anthropic files IPO prospectus, with risk disclosures taking center stage26/10/26 · Anthropic|Robot policy trained only in simulation matches top real-world systems26/10/26|OpenAI unveils GPT-6.1 Terra, a model built for computer use26/10/26 · OpenAI|New benchmark checks whether LLMs actually test their temporal medical claims29/10/26|An ODE-based sampling method speeds up discrete diffusion models29/10/26|OpenAI revises its Model Spec with five guiding values for AI behavior29/10/26 · OpenAI|Inside Proton's internal clash over its AI pivot29/10/26 · Proton|Mistral Makes Large 4 Available as Open Weights29/10/26 · Mistral AI|Gemini 4 Carbon tops the Artificial Analysis Intelligence Index v529/10/26 · Google|Mistral Large 4 now available on Hugging Face29/10/26 · Mistral AI|Google Introduces "Gemini 4 Carbon" as Next Step in Its Model Lineup28/10/26 · Google|DeepMind unveils Genie 4, a new generation of world model27/10/26 · Google DeepMind|Anthropic files IPO prospectus, with risk disclosures taking center stage26/10/26 · Anthropic|Robot policy trained only in simulation matches top real-world systems26/10/26|OpenAI unveils GPT-6.1 Terra, a model built for computer use26/10/26 · OpenAI|New benchmark checks whether LLMs actually test their temporal medical claims29/10/26|An ODE-based sampling method speeds up discrete diffusion models29/10/26|OpenAI revises its Model Spec with five guiding values for AI behavior29/10/26 · OpenAI|Inside Proton's internal clash over its AI pivot29/10/26 · Proton|
AI tools

Run your AIs

189 scored tools, ranked by overall score. Last checked on 29 October 2026. Refine with the Finder →

The three to remember

Most powerful
Ollama
AppSecurity: nothing to flag
9.7

The de facto standard for running models locally and connecting them to other tools.

  • · Utility 9.5/10, best of the selection
  • · Free, no limits
  • · No known security caveat
Free1 clickSite ↗
Simplest
Frontend Design (Anthropic)
SkillRisingSecurity: nothing to flag
9.6

The most installed skill on the official marketplace: one click, immediate effect on UI quality.

  • · Utility 9/10
  • · Free, no limits
  • · No known security caveat
Free1 clickSite ↗
The alternative
Context7
MCPSecurity: nothing to flag
9.4

The most installed MCP, and for good reason: it eliminates hallucinated APIs.

  • · Utility 9/10
  • · Free tier to get started
  • · No known security caveat
Freemium1 clickSite ↗
ToolPriceScoreSec.
OllamaRun an open-weight model in one command, with a local OpenAI-compatible API.Free9.7Security: nothing to flag
Ultimate Vocal RemoverRisingVoice and instrument separation running locally with multiple models (MDX, Demucs).Free9.6Security: nothing to flag
LangfuseRisingOpen source LLM observability: traces, costs, evaluations, prompt management.Freemium9.3Security: nothing to flag
mcp-scanRisingScans your MCP servers and installed skills for tool poisoning, rug pulls, and injections.Free9.3Security: nothing to flag
JanRisingOpen-source alternative to LM Studio, with zero telemetry.Free9.2Security: nothing to flag
spec-kit (GitHub)Spec-driven development toolkit: constitution, spec, plan, tasks, then implementation.Free9.2Security: nothing to flag
Buzz (Whisper)RisingLocal transcription with Whisper, SRT and VTT export, no account or cloud required.Free9.1Security: nothing to flag
Hermes Agent (Nous Research)RisingSelf-hosted agent that learns from its tasks, creates its own skills, maintains persistent memory, and responds on Telegram, Discord, Slack, WhatsApp, or Signal.Freemium9.1Security: nothing to flag
UpscaylRisingEnlarges and restores images up to 16x locally, without watermark.Free9.1Security: nothing to flag
vLLMRisingHigh-performance serving engine for running LLMs for multiple users in parallel.Free9.1Security: nothing to flag
MeetilyMeeting assistant that records, transcribes (Whisper or Parakeet) and summarises your meetings on your machine, with no bot and no cloud.Freemium9.0Security: nothing to flag
PromptfooRisingPrompt, agent, and RAG testing and red-teaming, CI-integrable.Freemium9.0Security: nothing to flag
REAL Video EnhancerRisingFree interface to interpolate and upscale videos using RIFE and ESRGAN.Free9.0Security: nothing to flag
Arize PhoenixRisingLLM application tracing and evaluation, built on OpenTelemetry.Freemium8.9Security: nothing to flag
ccusageRisingReads local Claude Code logs and displays cost by day, session or model.Free8.9Security: nothing to flag
garakRisingNVIDIA's LLM vulnerability scanner: jailbreaks, leaks, hallucinations, injections.Free8.9Security: nothing to flag
llama.cppRisingThe C/C++ inference engine behind Ollama, LM Studio, and most local AI tools.Free8.9Security: nothing to flag
CrewAIRisingPython framework for orchestrating teams of agents by roles.Freemium8.8Security: nothing to flag
LangGraphFramework for building stateful agents with an explicit control graph.Freemium8.8Security: nothing to flag
Cisco mcp-scannerRisingOpen source scanner by Cisco based on YARA to analyze MCP tool descriptions and schemas.Free8.7Security: nothing to flag
DSPyRisingFramework that replaces prompt writing with programming and automatic optimization.Free8.7Security: nothing to flag
HeliconeRisingObservability proxy: costs, latency and logs of all your LLM requests by changing one URL.Freemium8.7Security: nothing to flag
Home AssistantOpen source home automation with local voice assistant Assist.Free8.7Security: nothing to flag
SGLangRisingA serving engine competing with vLLM, very fast thanks to the RadixAttention prefix cache.Free8.7Security: nothing to flag
ToolHiveRisingRuns each MCP server in an isolated container with managed permissions and secrets.Free8.7Security: nothing to flag
PortkeyRisingAI gateway with routing, guardrails, and observability, open source since March 2026.Freemium8.6Security: nothing to flag
NeMo GuardrailsRisingNVIDIA's programmable guardrails to control dialogues, topics, and tool calls.Free8.5Security: nothing to flag
BifrostRisingAI gateway in Go, very low latency, also serving as an MCP gateway.Freemium8.4Security: nothing to flag
Guardrails AIRisingLLM input and output validators (PII, toxicity, format) with a validator hub.Freemium8.4Security: nothing to flag
Wan2GPRisingInterface to run Wan, Hunyuan and LTX on modest GPUs.Free8.4Security: nothing to flag
ACE-Step 1.5RisingFull music generation (vocals, structure, style) running locally, open source.Free8.3Security: nothing to flag
LocalAIRisingSelf-hosted OpenAI-compatible API covering text, image, audio and embeddings.Free8.3Security: nothing to flag
MacWhisperLocal Whisper transcription on Mac, with a polished interface.Freemium · from 59 €8.3Security: nothing to flag
MLX (mlx-lm)RisingApple framework for running LLMs on M-series chips using unified memory.Free8.3Security: nothing to flag
PyRITRisingMicrosoft's AI red-teaming toolkit.Free8.3Security: nothing to flag
Applio (RVC)RisingModern interface for RVC voice conversion, training and inference running locally.Free8.2Security: one caveat, see the card
FrigateRisingLocal NVR for cameras with 100% local AI object detection.Freemium8.2Security: nothing to flag
KoboldCppRisingSingle executable to run GGUF models with an interface oriented towards writing and roleplay.Free8.2Security: nothing to flag
Claude Code Usage MonitorRisingReal-time monitor for Claude Code quotas and tokens in the terminal.Free8.1Security: nothing to flag
ExaSearch engine designed for agents, with an official MCP integration.Freemium8.1Security: nothing to flag
LangSmithObservability and evaluation platform from LangChain, usable outside LangChain.Freemium8.1Security: nothing to flag
TRELLIS.2 (Microsoft)RisingMicrosoft image-to-3D model producing a PBR mesh in seconds, running locally.Free8.1Security: nothing to flag
Open WebUIRisingSelf-hosted chat interface in the style of ChatGPT, connected to Ollama or any OpenAI API.Free8.0Security: one caveat, see the card
AutoGenRisingMicrosoft's multi-agent framework, a pioneer in the field.Free7.9Security: nothing to flag
BraintrustEvaluation and logging platform for AI products, oriented towards product teams.Freemium7.9Security: nothing to flag
Claude SquadRisingManages multiple terminal agents (Claude Code, Codex, OpenCode, Amp) in isolated workspaces.Free7.9Security: one caveat, see the card
FirecrawlRisingTurns any website into clean markdown for AI, with an MCP server.Freemium7.9Security: one caveat, see the card
InvokeAIRisingLocal generation studio with a retouching canvas, more accessible than ComfyUI.Freemium7.9Security: nothing to flag
TavilySearch and extraction API optimized for LLMs, with MCP support.Freemium7.9Security: nothing to flag
Video2XWatch outLocal video upscaling and frame interpolation (Real-ESRGAN, RIFE), open source.Free7.9Security: nothing to flag
ComfyUIRisingA node-based workshop for running open-source image and video models locally.Free7.8Security: one caveat, see the card
Google Agent Development Kit (ADK)NewGoogle's open-source kit for designing multi-agent systems in Python, Java, Go, or TypeScript, with the A2A protocol and direct deployment to Vertex AI.Free7.8Security: one caveat, see the card
LiteLLMSelf-hosted gateway that unifies 100+ providers behind an OpenAI API, with budgets and virtual keys.Freemium7.8Security: one caveat, see the card
Microsoft Agent FrameworkNewMicrosoft open-source framework for .NET and Python, successor to Semantic Kernel and AutoGen: agents, graph-based workflows, A2A and MCP.Free7.8Security: one caveat, see the card
OpenAI Agents SDKNewOpenAI's open source framework for building agents and agent teams: handoffs between agents, guardrails, sessions, and built-in tracing.Free7.8Security: one caveat, see the card
FooocusWatch outUltra-simplified local interface to generate high-quality SDXL images without technical adjustments.Free7.7Security: nothing to flag
GPT-SoVITSVoice cloning and synthesis from a few seconds of audio, running locally.Free7.7Security: one caveat, see the card
RequestyHosted router with cost optimization and guardrails, team-oriented.Paid7.7Security: nothing to flag
StrixOpen-source AI agent pentesting: agents run your application, find vulnerabilities, and prove each one with a PoC.Freemium7.7Security: one caveat, see the card
TripoSRImage to 3D in under one second, open source, runs locally.Free7.7Security: nothing to flag
AnythingLLMRisingAll-in-one application to chat with your documents, locally or via API, with built-in agents.Freemium7.6Security: one caveat, see the card
Mem0RisingPersistent memory layer for agents and assistants, hosted or self-hosted.Freemium7.6Security: one caveat, see the card
LM StudioA graphical interface to download and chat with open-weight models on your own PC.Freemium7.5Security: one caveat, see the card
Postshot (Jawset)Watch outLocal Gaussian splat training on NVIDIA GPU.Freemium7.5Security: nothing to flag
Not DiamondSmart router that selects the cheapest model capable of handling each request.Freemium7.4Security: nothing to flag
OpenClawRisingOpen-source self-hosted personal agent that acts on your behalf: emails, calendar, files, commands, browser, controlled from your messaging apps.Free7.4Security: one caveat, see the card
OpenRouterOne API key to access 400+ models, with automatic fallback between providers.Freemium7.4Security: one caveat, see the card
whisper.cppRisingC/C++ implementation of Whisper, very fast on CPU and Apple Silicon.Free7.4Security: nothing to flag
Zep / GraphitiRisingTemporal knowledge graph memory for agents, with Graphiti in open source.Freemium7.4Security: one caveat, see the card
F5-TTSRisingSpeech synthesis and voice cloning via flow matching, fast and local.Free7.3Security: one caveat, see the card
FramePackWatch outLocal long video generation with only 6 GB of VRAM.Free7.3Security: nothing to flag
herdrAgent multiplexer in Rust that detects Claude Code, Codex or Amp and displays their status: blocked, working, done, idle.Free7.3Security: one caveat, see the card
LivePortraitAnimates a portrait from a reference video, locally.Free7.3Security: one caveat, see the card
cmuxRisingNative macOS terminal built on Ghostty, designed for agents: vertical tabs, notifications, built-in browser.Free7.2Security: one caveat, see the card
LettaRisingManaged-memory agent platform, successor to MemGPT.Freemium7.2Security: one caveat, see the card
Vibe KanbanRisingKanban board to dispatch tasks to coding agents and track their progress.Free7.1Security: one caveat, see the card
YuERisingGenerate complete songs with lyrics, locally, open source.Free7.1Security: one caveat, see the card
ConductormacOS app that runs multiple Claude Code agents in parallel with a dashboard.Freemium7.0Security: one caveat, see the card
GPT4AllWatch outLocal public chat client by Nomic, with document discussion support.Free7.0Security: nothing to flag
Strands AgentsNewAWS open source SDK for writing an AI agent in a few lines of Python or TypeScript: the model drives the loop, you declare the tools.Free7.0Security: one caveat, see the card
Invariant GuardrailsWatch outSecurity rules applied in real time to agent tool calls.Freemium6.8Security: nothing to flag
MstyA polished desktop client that combines local models and cloud APIs in a single interface.Freemium6.7Security: one caveat, see the card
text-generation-webuiRisingHistoric web interface for local LLMs, highly configurable.Free6.7Security: one caveat, see the card
DeepFilterNetWatch outReal-time noise suppression, lightweight, works even on CPU.Free6.6Security: nothing to flag
Hunyuan3D 2.1 (Tencent)Watch outTextured PBR 3D generation from an image, running locally.Free6.6Security: one caveat, see the card
Real-ESRGANWatch outLocal image and video upscaling, the backbone of many tools.Free6.6Security: nothing to flag
RIFEWatch outAI-based video frame interpolation, running locally.Free6.6Security: nothing to flag
Amazon Bedrock AgentCoreNewAWS managed runtime to put an agent into production: per-session isolated execution, memory, tool gateway, identity, browser, code sandbox, and traces.Paid6.5Security: one caveat, see the card
CrystalWatch outOpen source desktop app to launch and compare multiple Claude Code sessions in worktrees.Free6.4Security: one caveat, see the card
Mira-SceneNewReconstructs an editable 3D scene from a single photo: each object is modelled separately then placed back at its position and scale.Free6.4Security: one caveat, see the card
Resemble EnhanceWatch outSpeech denoising and enhancement, open source.Free6.4Security: nothing to flag
MCP DefenderDesktop app that intercepts and filters MCP traffic from installed AI clients.Freemium6.3Security: one caveat, see the card
QdrantNewOpen source vector database written in Rust: it stores your documents as vectors and retrieves the closest passages by meaning, the search building block of a RAG system.Freemium6.2Security: one caveat, see the card
DiffRhythmWatch outGenerate complete tracks in seconds, locally.Free6.1Security: one caveat, see the card
Vertex AI Agent EngineNewGoogle Cloud managed runtime for deploying agents: execution, sessions, long-term memory (Memory Bank), code execution, and agent identity.Paid6.1Security: one caveat, see the card
LatentSync (ByteDance)Watch outHigh-quality diffusion-based lip sync, running locally.Free6.0Security: one caveat, see the card
MuseTalkWatch outReal-time lip-sync running locally.Free6.0Security: one caveat, see the card
MusicGen (Meta)Watch outMeta's text-to-music generation model with open weights.Free5.9Security: one caveat, see the card
OmniRouteRisingLocal open-source gateway in front of 290+ providers with automatic quota-based switching.Free5.5Security: serious reservations

App : A product you use directly · Tool : A precise action the AI can call · MCP : Standardises the connection between an AI application and systems that expose tools or data · Built-in mode : A feature to switch on inside an assistant you already use (ChatGPT, Claude, Gemini): nothing to install · Skill : A reusable method: instructions, procedures, sometimes scripts or templates · Plugin : A set of capabilities installed into an AI application or a host

Method

Overall score = utility (rated by the editorial team), reliability (GitHub activity, official publisher, reachable site), security (where data goes, permissions, known incidents) and accessibility (install, required level, price), weighted per category. A tool inactive for over six months is flagged and leaves the recommendations. No ranking is sold.

Missing a tool? A score looks wrong?

Write to the editors