The momentum of Kong Insomnia continues with the latest v10 GA release, which ships with an incredible amount of capabilities like the long-awaited Collection Runner built into the application, a new governance capability called “Invite Control,” and an entirely new capability for AI consumption.
Let’s look at these features one by one.
Insomnia v10 ships with a built-in Collection Runner that allows you to run an unlimited number of collection runs whether you are storing your data locally, on Git, or in the cloud.
By using the Collection Runner you can easily test and verify that your collections are always running properly and that the APIs you're consuming aren't changing. By validating that the requests and responses are always accurate, you can get a higher confidence level that your APIs keep working over time. It also helps with finding problems more quickly in case one of the endpoints changes, inadvertently breaking your client implementations.
The Collection Runner also allows you to order the execution of the requests, to run multiple iterations of your tests, and to upload custom CSV or JSON data that can be used in your pre-request and after-response scripts, very useful when testing with sample data for example.
Of course, with the new “Test Results” section, you can now also run tests for an individual request without having to use the Collection Runner.
You can run individual requests and see the test outcomes in the new “Test Results” section.
The new Collection Runner is available in all Insomnia subscriptions, including the free plan.
With the new Invite Control enterprise capability, you can now determine what domains are allowed to be invited to work on your organizations and projects. This ensures that external users aren't mistakenly invited.
While the decision to invite someone is still 100% in the control of the organization’s admins, setting up Invite Control rules further ensures that only approved domains can be added to your projects, collections, and design specs. For example, if an admin has mistakenly invited someone that they shouldn’t have (such as a user with a personal email address instead of a corporate email), Invite Control can catch the error immediately and prevent the invite from being sent.
With Invite Control we introduce even better governance and security with our collaborators.
The capability to invite users to an organization can be entirely disabled with the Invite Control feature as well.
This capability, in addition to Storage Control and the many other enterprise governance, compliance, and security capabilities in Insomnia, helps to ensure that your API assets are always secure and accessible only by authorized individuals in your organization.
Finally, we're introducing a new breakthrough AI product that has never been seen before in the world, our new Insomnia AI Runner which delivers intelligent multi-LLM infrastructure for AI to every Insomnia developer.
Insomnia’s new AI Runner offering is the easiest way to build production GenAI applications with multi-LLM support, semantic caching, and AI traffic analytics collection.
With the new Insomnia AI Runner, millions of Insomnia users can now consume multiple GenAI LLMs by writing their code only once. Out of the box, users can redirect their applications to use Azure AI, AWS Bedrock, GCP Vertex, OpenAI, Anthropic, Mistral, and Cohere without any code changes. This accelerates experimentation, benchmarking, and the final implementations of GenAI applications and agents.
Also, we are making it very easy to set up guardrails for improving GenAI security for content moderation, prompt jailbreaking, prompt injection, and a lot more.
Setting up caching preferences and security guardrails has never been simpler.
We're also introducing intelligent semantic caching capabilities to the Insomnia AI Runner, enabling you to accelerate GenAI requests by up to 20x. Insomnia users can also significantly reduce costs by processing the underlying LLM consumption with a globally distributed cache for GenAI requests that can semantically understand the meaning of your prompts, even when different words are being used. This allows Insomnia users to build better — faster — AI experiences in all of your digital products.
Build once and use all the LLMs supported by AI Runner with semantic caching for better performance.
You can also optionally decide to use client credentials sent by your code or to store those credentials in the AI Runner itself using Kong’s secure enterprise cloud for centralized management of credentials.
Powered by Kong’s enterprise-grade Konnect platform and AI Gateway solution, this offering is designed for compliance with SOC2, CSA Start Level 1, NIST 800-218, and more.
The underlying technology that powers Insomnia AI Runner can also be self-hosted on Konnect or on-premises. To learn more about running your very own cluster of the Kong AI Gateway please reach out to the Kong sales team.
For more information on the new AI Runner, read the full blog post here.
To start using the new capabilities, get started for free with Insomnia 10 today.
As of September 2025, Kong Gateway Enterprise 3.8 will enter its End Of Life (EOL) phase and will no longer be fully supported by Kong. Following this, Kong Gateway Enterprise 3.8 will enter a 12-month sunset support period, focused on helping cus
We're very excited to announce Kong Mesh 2.12 to the world! Kong Mesh 2.12 delivers two very important features: SPIFFE / SPIRE support, which provides enterprise-class workload identity and trust models for your mesh, as well as a consistent Kuma R
Ever feel like you're fighting an uphill battle with your API strategy? You're building APIs faster than ever, but somehow everything feels harder. Wasn’t API-first supposed to make all this easier? Well, you're not alone. And now industry analys
It’s been almost a year since we released our Konnect Terraform provider . In that time we’ve seen over 300,000 installs, have 1.7 times as many resources available, and have expanded the provider to include data sources to enable federated managem
Kong-quer the Agentic AI Hackathon 🚀 Calling all builders, tinkerers, and API innovators. The Kong Hackathon is back for API Summit 2025 ! This year, we’re challenging developers worldwide to create projects that don’t just react, they think , a
In my previous post, we discussed how we can implement a basic AI Agent with Kong AI Gateway. In part two of this series, we're going to review LangGraph fundamentals, rewrite the AI Agent and explore how Kong AI Gateway can be used to protect an LLM