Alternatives to Patronus AI

Polymath builds simulation environments for training and evaluating autonomous AI agents on long-horizon tasks

Automated AI red teaming that discovers and exploits vulnerabilities in production AI agents and models

Arena AI is a free, community-voted LLM leaderboard ranking chat models, agents, and fullstack code on live head-to-head battles.

OpenCompass (司南) benchmarks LLMs, VLMs, and AI4S models with 100+ open evaluation datasets

Open-source benchmark for AI agents on multi-step, real-world software company tasks.

AppWorld is a simulated-world benchmark for evaluating AI coding agents across 9 apps and 457 APIs.

Silico is Goodfire's interpretability agent for understanding, debugging, and controlling the internals of your AI models

Weights & Biases tracks ML experiments and traces LLM apps so teams can ship AI models faster

Agentic law platform that embeds legal judgment into AI agents for verifiable compliance

Hume AI provides real human feedback, simulation, and expression measurement for emotionally intelligent voice AI.

Paste one prompt, see how 40+ AI models answer it, then pick the one that's best, fastest, or cheapest.

ViDoRe is an open visual document retrieval benchmark and ColPali-style model suite for enterprise RAG evaluation.

A Carnegie Mellon research benchmark of 910 visually grounded web tasks for multimodal browser agents, scored by execution rather than string matching.

Open-source benchmark that tests AI agents on real, live websites with two-stage HTTP interception and LLM-judge scoring.

Free public leaderboard ranking LLMs on real-world agentic tasks — planning, tool use, and multi-step execution

VLMEvalKit benchmarks 220+ vision-language models across 80+ multimodal tasks with an open leaderboard.

ECC is an open-source agent harness layer that hardens and optimizes coding agents across Claude Code, Codex, Cursor, OpenCode and Kimi.

Andon Labs runs real cafés, stores, and radio stations with AI agents in charge, then publishes what breaks.

The AI Incident Database (AIID) is a free, open archive of real-world AI harms you can search and download.

Open-source observability and self-repair for AI agents, turning production failures into validated, reusable rules.

Iris.ai builds an AI knowledge foundation that turns complex regulated enterprise data into auditable, explainable intelligence.

Imbue is an open AI lab building coding agent tools that run in parallel and answer to you, not a vendor.

Fiddler AI is an enterprise AI control plane for agent observability, guardrails, and governance across the agentic lifecycle.

Floutwork consolidates GPT, Claude, Gemini and Grok into one governed AI workspace with shared memory, AI assistants and admin controls for teams of 10–1,000.

Independent AI leaderboard ranking 398 LLMs by a composite LLM Stats Score across intelligence, speed, and price.

Open ICLR-published benchmark measuring agentic LLM and VLM reasoning across seven procedurally generated games, with weekly-updated public leaderboards.

Free benchmark leaderboard that scores LLMs on elite competition math from AIME and IMO to Lean proof datasets

Open-source multi-agent offensive-security harness that turns the AI coding agent already on your machine into an autonomous vulnerability hunter