Overview
This is paid repackaged open-source software supplied by Software Sushi. DeepEval 4.2.4 core uses Apache License 2.0; dependencies retain their own licenses and original notices. The image supplies the unmodified upstream DeepEval CLI, a pinned Python environment and local evaluation examples on Ubuntu 24.04 x86_64. Software Sushi is not affiliated with or endorsed by Confident AI. Confident AI hosted services, dashboard subscriptions and commercial platform features are not included.
The supplied example evaluates two fixed synthetic returns-policy answers. The correct answer states a 30-day window with a receipt; the incorrect answer states 90 days without one. DeepEval JSON correctness checks run locally without a model. An optional configured OpenAI-compatible model acts as a separate GEval judge of those answers. It does not generate the assistant answers. Reports retain original scores, thresholds, reasons and good/bad results, then a separate reader verifies the saved example. Model scores can vary and need human review.
Start with t3.medium, 4 GiB RAM and at least 24 GiB root storage with EBS encryption enabled. No GPU is required for the examples. Connect through AWS Systems Manager Session Manager with your own appropriate instance role and outbound endpoint access, or use your own SSH key and restricted network setup. The product has no web UI, HTTP API listener or application inbound port. Access is controlled by the Ubuntu operating-system account.
Run sudo -iu ubuntu bash /opt/sushi/docs/demo.sh. To add a judge, run sushi-deepeval configure as ubuntu and enter an entitled provider endpoint, exact model ID and provider key through the hidden prompt. Then run bash /opt/sushi/docs/demo.sh --judge. Remote endpoints require HTTPS and JSON-object chat responses. No provider key, model weights, provider credits or Confident AI account is supplied. Judge requests send the synthetic question, the answer under test and the reference answer to your chosen endpoint; review provider privacy and retention terms before using sensitive data.
Each run creates a new JSON report under /home/ubuntu/deepeval-results. Settings under /home/ubuntu/.config/deepeval/judge.json use mode 0600. Reports and settings persist through reboot on the same root volume. Back them up before replacement or termination. The on-instance README explains configuration, native CLI use, independent readback and recovery. Original source distributions, wheels and notices are retained on the instance. Additional upstream workflows can require optional packages or separate service agreements.
Software is USD 0.03 per running instance-hour on each offered type. EC2, EBS, data transfer and model-provider usage are billed separately. English getting-started guide: https://softwaresushi-marketplace-branding-public.s3.us-east-1.amazonaws.com/docs/deepeval/9a7b75dab712492d745e/getting-started.html
Highlights
- Run DeepEval local JSON correctness checks against explicit good and bad synthetic answers.
- Optionally use your entitled OpenAI-compatible model as a separate GEval factual-correctness judge.
- Keep original JSON reports, scores and thresholds on the instance with independent saved-result readback.
Details
Features and programs
Financing for AWS Marketplace purchases
Pricing
Free trial
Dimension | Cost/hour |
|---|---|
t3.medium Recommended | $0.03 |
t3.large | $0.03 |
t3.xlarge | $0.03 |
t3.2xlarge | $0.03 |
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 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
Packaging candidate 0.1.1; source 18c21574619a. Automatic EC2 boot, real use and host reboot verified on this exact AMI. Known OS/kernel High/Critical vulnerability findings remain unresolved; no security fix or security qualification is claimed.
Additional details
Usage instructions
Launch t3.medium or a larger supported instance with at least 4 GiB RAM and 24 GiB encrypted gp3 root storage. Require IMDSv2 and an instance profile with AmazonSSMManagedInstanceCore. No inbound rule or key pair is needed for Session Manager. Connect with Session Manager, then run sudo -iu ubuntu. Read /opt/sushi/docs/README.md and run bash /opt/sushi/docs/demo.sh. Keep application ports closed; use the loopback access and tunnel instructions in the installed guide. Back up your data before replacing the instance. EC2, EBS and data transfer charges apply separately.
Resources
Vendor resources
Support
Vendor support
Software Sushi provides AMI packaging and setup guidance. Contact info@softwaresushi.com for installation questions. Support coverage requires a separate agreement; upstream project and model-provider support are not included.
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.