Listing Thumbnail

    Trinity

     Info
    Sold by: Ability AI 
    Deployed on AWS
    Compounding intelligence your company owns. Trinity is the operating system for the AI-native company: persistent AI agents that run your company's work, keep their memory, share what they learn, and ship their own improvements, all inside your own AWS account. Every agent's state is versioned in git and every action is recorded in an append-only audit log, with human approval gates where you need them. It works with any model (Claude, OpenAI, Gemini, or open weights) and is open source under Apache 2.0. Your infrastructure, your data, your agents: export everything and walk away whenever you like.

    Overview

    Play video

    Most AI tools are rented. Whatever they build up about your business lives on someone else's platform, and it starts over when you switch vendors. Trinity is built the other way round. It is the operating system where an AI-native company's agents and workflows run with permissions, state, governance and receipts, deployed in your own AWS account so that the intelligence it builds up belongs to you.

    What Trinity does

    • Persistent agents. Each agent runs in its own isolated container with its own filesystem, credentials and memory. It runs around the clock and keeps its context across restarts.
    • Works on its own. Agents run on cron schedules, webhooks and event triggers, so work gets done whether or not anyone is at a keyboard.
    • Reachable from anywhere. Use the web UI or REST API, or reach agents through Slack, Telegram, WhatsApp and browser voice.
    • Multi-agent orchestration. Orchestrator/worker patterns, parallel fan-out, shared folders and event pub/sub between agents.
    • Shared, not siloed. Your whole team works with the same agent and its growing context, with four-tier role-based access (admin, creator, operator, user).
    • Observability. Per-agent execution history, cost tracking, context usage, error rates and a fleet health dashboard.
    • Guardrails. Per-agent timeouts, capacity limits, read-only mode and a human-in-the-loop approval queue.

    Receipts, not promises Every agent's state is version-controlled in git, so you can roll back to any point. The audit log is append-only, keeps 365 days and records every tool call. A SHA-256 hash chain is available for compliance mode, and you enable it per install. Credentials are encrypted with AES-256-GCM and injected at runtime.

    Yours to keep

    • Single-tenant. It runs in your VPC and your data stays in your network.
    • Model-agnostic. Use Claude, OpenAI, Gemini or open-weight models with your own keys.
    • Open source under Apache 2.0, free for commercial use and self-hosting. Export agent definitions, skills and memory at any time.

    Trinity Enterprise For teams that need more governance, Trinity Enterprise adds user management, export of the audit log to your SIEM, and an audit dashboard. It's licensed separately and held by you. Contact hello@ability.ai .

    About Ability AI Ability AI is the lab for AI-native companies, and the lab behind Trinity and Cornelius, its cognitive architecture for making a company's work compound. We run our own company on Trinity.

    Highlights

    • Compounds, and you own it. Agents keep their memory and context across restarts. What they learn builds up as an asset in your AWS account, not on a vendor's platform.
    • Governed, with receipts. Agent state is versioned in git, the audit log is append-only with 365-day retention and records every tool call, and approval queues put a human gate on the actions that matter.
    • Runs in your account, on any model. It's single-tenant and inside your VPC, and works with Claude, OpenAI, Gemini, or open-weight models. Apache 2.0 open source with full export, so there's no lock-in.

    Details

    Delivery method

    Delivery option
    64-bit (x86) Amazon Machine Image (AMI)

    Latest version

    Operating system
    Ubuntu 24.04 LTS

    Deployed on AWS
    New

    Introducing multi-product solutions

    You can now purchase comprehensive solutions tailored to use cases and industries.

    Multi-product solutions

    Features and programs

    Financing for AWS Marketplace purchases

    AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
    Financing for AWS Marketplace purchases

    Pricing

    This product is available free of charge. Free subscriptions have no end date and may be canceled any time.
    Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator  to estimate your infrastructure costs.

    AI Insights

     Info

    Dimensions summary

    Trinity software is free. You pay only for the AWS EC2 instance you run it on, billed per hour. You pick from nine instance types in two sizes. The large sizes (m5.large, m6i.large, m7a.large, t3.large, t3a.large) carry fewer compute resources. The xlarge sizes (m5.xlarge, m6i.xlarge, m7a.xlarge, t3.xlarge) carry more. The m-series and t-series families differ in processor generation and burst behavior. Hourly cost rises with the instance you select, so you match the size to your workload. No separate license fee applies.

    Top-of-mind questions for buyers

    The large sizes carry fewer vCPUs and less memory than the xlarge sizes. Trinity's own guidance suggests about 4 vCPU and 8 GB for the platform plus its supporting services. The xlarge options give more headroom when you run several agents at once.
    Hourly charges meter running instance-hours only. A fully stopped instance does not accrue hourly compute charges. Stopped instances may still incur AWS storage fees for attached volumes. The Trinity software itself carries no license fee at any state.
    The m-series instances deliver steady, fixed compute each hour. The t-series instances use burstable performance, giving baseline compute with short bursts. Both bill per hour. Choose t-series for intermittent agent runs and m-series for continuous, steady workloads.
    www.ability.ai
    Helpful?

    Vendor refund policy

    Trinity is free. Ability AI charges nothing; AWS bills the EC2, EBS and IPv4 usage directly.

    How can we make this page better?

    Tell us how we can improve this page, or report an issue with this product.
    Tell us how we can improve this page, or report an issue with this product.

    Legal

    Vendor terms and conditions

    Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .

    Content disclaimer

    Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.

    Usage information

     Info

    Delivery details

    64-bit (x86) Amazon Machine Image (AMI)

    Amazon Machine Image (AMI)

    An AMI is a virtual image that provides the information required to launch an instance. Amazon EC2 (Elastic Compute Cloud) instances are virtual servers on which you can run your applications and workloads, offering varying combinations of CPU, memory, storage, and networking resources. You can launch as many instances from as many different AMIs as you need.

    Version release notes

    First AWS Marketplace release of Trinity. HTTPS on the instance's public IP with no domain required, admin setup gated on the instance ID, bundled PostgreSQL, and automatic certificate renewal after a stop/start or an Elastic IP change.

    Additional details

    Usage instructions

    1. Launch the instance with a public IPv4 address and the recommended security group (ports 443 and 80). First boot takes a few minutes.
    2. In the EC2 console, open Instances and copy this instance's Instance ID (it starts with i-) and its Public IPv4 address.
    3. Open https://<public IPv4 address>/. Enter the instance ID when asked, then choose an admin password. An email address is optional; without one, sign in as admin with that password.
    4. Add a model credential in the browser and create your first agent. The first-run guide walks through attaching a domain and a Cloudflare Tunnel when you are ready to move off the bare IP.

    Stopping and starting: without an Elastic IP, AWS gives the instance a new public address after a stop/start. Trinity picks it up and obtains a new certificate within about five minutes; open the new address after that. An Elastic IP keeps the address fixed.

    Managing over SSH (optional, open port 22 from your own address only): cd /opt/trinity, then docker compose -f docker-compose.hosted.yml ps or logs -f backend, and sudo ./scripts/deploy/start.sh --hosted to restart. The database is a bundled PostgreSQL container. Backups run nightly under /opt/trinity/trinity-data/backups/; take EBS snapshots as well.

    Support: https://github.com/abilityai/trinity/issues  (label aws-marketplace). Docs: https://docs.ability.ai 

    Resources

    Vendor resources

    Support

    Vendor support

    AWS infrastructure support

    AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.

    Similar products

    Customer reviews

    Ratings and reviews

     Info
    0 ratings
    5 star
    4 star
    3 star
    2 star
    1 star
    0%
    0%
    0%
    0%
    0%
    0 reviews
    No customer reviews yet
    Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.