Overview
Hermes - Software Sushi product illustration
Original AI-generated illustration by Software Sushi. Representative product artwork; not an application screenshot or an upstream official logo.
Hermes - Software Sushi product illustration
Hermes AMI Setup: Telegram Agent Walkthrough
Hermes AI Agent turns an EC2 instance into a self-hosted AI assistant that customers can operate through Telegram. This is a repackaged open source software product wherein additional charges apply for independent AWS packaging, validation, security hardening, release automation, and support. Hermes Agent is developed by Nous Research and distributed under the MIT license; this product does not claim authorship of the upstream project.
The production-ready AMI includes a pinned, published Hermes Agent release, its Telegram gateway dependencies, UFW default-deny inbound firewalling, fail2ban SSH protection, SSH key-only authentication, unattended security upgrades, network sysctl hardening, gp3 root storage with encryption available at launch, and required IMDSv2. The gateway runs as a systemd service with customer configuration and state stored under /home/ubuntu/.hermes. No customer credentials are baked into the AMI.
Getting started after launch:
- Restrict TCP 22 to a trusted CIDR and connect as ubuntu with your EC2 key pair.
- Set your model-provider API key, model selection, Telegram bot token, and allowed Telegram user ID.
- Run sudo /opt/hermes/bin/provision-hermes.sh and verify systemctl status hermes-gateway.
The product is fully functional after configuration and has no product-imposed time or usage restriction. An internet connection, a supported third-party model-provider account, and a Telegram bot are required and are disclosed external dependencies. The optional five-day free trial changes only the software billing and automatically converts to the paid hourly offer; it does not enable a separate trial or evaluation edition.
The software charge is USD 0.02 per running instance-hour. EC2, EBS, data transfer, Telegram connectivity, and third-party model-provider usage are charged separately by their providers.
Highlights
- Ready to configure: Hermes Agent and its Telegram gateway dependencies are pinned and pre-installed; add your own model-provider and Telegram credentials after launch.
- Production-operated foundation: Hermes runs as a systemd service with release markers, persistent customer state, automated security updates, and daily Marketplace-copy health checks.
- Hardened from first boot: UFW, fail2ban, key-only SSH, gp3 storage with encryption available at launch, unattended security upgrades, and required IMDSv2 reduce the host attack surface.
Details
Features and programs
Financing for AWS Marketplace purchases
Pricing
Free trial
Dimension | Cost/hour |
|---|---|
t3.medium Recommended | $0.02 |
t3.large | $0.02 |
t3.small | $0.02 |
t3.xlarge | $0.02 |
m6i.xlarge | $0.02 |
m6i.large | $0.02 |
Dimensions summary
Top-of-mind questions for buyers
Vendor refund policy
We offer a full refund of software charges for the first 48 hours billing starts if the product does not perform as described. To request a refund, email info@softwaresushi.com with your EC2 instance ID and a description of the issue. AWS infrastructure costs (EC2, EBS, data transfer) are billed by AWS and are not eligible for refund by us.
How can we make this page better?
Legal
Vendor terms and conditions
Content disclaimer
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
Standalone Hermes AI Agent v0.23.1 packaged with the native PM-managed Python 3.14 runtime and preserved hermes, hermes-python and optional provision-hermes.sh interfaces. Pillow 13.0.0 is built from its pinned source distribution against patched Ubuntu Brotli instead of using the prebuilt wheel's bundled library. Configure customer provider and channel credentials after launch. Back up customer state before migrating and do not restore old application trees, tools, installs or runtime caches. No messaging gateway is enabled by default.
Additional details
Usage instructions
Standalone Hermes AI Agent v0.23.1 on Ubuntu 24.04 LTS. Start with t3.medium (4 GiB RAM) and at least 20 GiB of encrypted root storage; larger customer workloads may need more resources. No application listener or messaging gateway is enabled by default.
- Launch this version with IMDSv2, outbound HTTPS and an instance role containing AmazonSSMManagedInstanceCore. Connect through AWS Systems Manager Session Manager; use sudo -iu ubuntu for application commands. No inbound application port is required; the recommended SSH rule is loopback-only. If you deliberately use SSH, restrict TCP 22 to your administrative CIDR.
- Wait for cloud-init to finish. In the session, run: sudo -iu ubuntu bash /opt/sushi/docs/demo.sh This checks the installed identity, native CLI and synthetic local file/terminal tools. It does not contact a model or Telegram.
- To configure your own provider interactively, run: sudo -iu ubuntu Then run: hermes setup Hermes uses its native package manager and Python runtime. The stable commands are hermes and hermes-python. Customer configuration and state belong under /home/ubuntu/.hermes.
- Optional compatibility setup remains available at /opt/hermes/bin/provision-hermes.sh. This helper is not run during image construction; when explicitly invoked it configures and starts a Telegram gateway. Supply its required environment through a root-owned mode-0600 file loaded locally after launch. Never put API keys or bot tokens in command arguments, user-data, shared logs or recordings. Keep the default allowlist policy, authorize only the intended Telegram user IDs, and ensure one polling owner per bot. Provider and messaging operation require your own credentials and configuration.
- Back up customer state before moving to a replacement AMI. Migrate only customer state and private configuration; do not copy an older tools directory, installs directory, runtime cache or application tree into this version. Browser, computer-use and gateway integrations are opt-in. Model-provider, EC2, EBS and data-transfer charges remain separate from the existing Marketplace software terms.
Resources
Vendor resources
Support
Vendor support
For AMI packaging, hardening, deployment, or billing support, email info@softwaresushi.com .
For upstream Hermes Agent issues, use this issue tracker: https://github.com/NousResearch/hermes-agent/issues
Support is provided in English during US business hours. Security reports should use the subject Hermes AI Agent security.
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.