MarkosWebWeb intelligence for the agentic economy

Olam Labs

ActiveAI-native

Building multi-agent simulations for model evals and training.

olamlabs.aiAI readiness report

What it does

Multi-agent environments let us evaluate and train models in complex, simulated worlds. We work with researchers on both evaluations and training for character and agentic performance typically difficult to assess for using current popular datasets. Play one of our first releases, Multi-Agent Arena (https://olamlabs.ai/arena), where humans come play social strategy games against multiple AI agents. It's currently top 50 on OpenRouter and has thousands of matches played each day. We use Multi-Agent Arena to build datasets on agentic performance and real-world socialization.

Its site says now · captured 2026-08-23

Olam Labs — Multi-Agent Arena

Olam Labs measures social intelligence in AI models for safety, agent performance, and real-world use, through behavioral evaluations in multi-agent simulations where models cooperate, deceive, and negotiate.

Next to it in Data for AI · synthetic data and model evals via simulations

  • PerfectBit, Inc.Y Combinator X26

    Correct by construction AI training data

  • QokedasY Combinator F26

    Data for AI Science

  • AbundantY Combinator F24

    Agent simulation and RL for researchers

  • DatacurveY Combinator W24

    Frontier coding data for training and evaluating LLMs

  • NeosyncY Combinator S22

    Neosync is an open-source anonymization and synthetic data platform.

  • MaingenY Combinator S26

    Simulating Agents Running Industrial Companies

  • LightwheelPlug and Play PnP 2026

    生成AIで事故等の希少データを無限生成し、ロボットや自動運転AIの学習を加速させる合成データ基盤

  • MarkovY Combinator S26

    Data for computer-use ai