TensorTest private evals
Private evaluation service for AI labs that runs an unreleased checkpoint or agent on the company's creative-editing jobs and returns a failure report, alongside RL environments packaged for a lab's sandbox and commissioned training and eval data; begins with video editing through the TimelineBench benchmark; terms are scoped per engagement and no prices are published.
Find alternatives to TensorTest private evals
What it does
- Evaluation and observability
- Data labelling
Features 4
- Private evalsRuns a lab's unreleased checkpoint or agent on the company's jobs and sends a failure report; results stay private.
- RL environmentsPackages the jobs for a lab's sandbox with the company's checks exposed as the reward.
- Training and eval dataNew jobs on commissioned footage with training rights, plus editors' trajectories, preference pairs and rubrics.
- TimelineBenchA benchmark that gave 16 agent setups 56 real editing jobs, each a whole job from raw footage and brief to finished cut.
Sources
- Pricing
-
No pricing page is published; the home page says "Terms are scoped per engagement." and offers Request a private eval.
- Description
-
Official documentation · 20 Oct 2026
Home page: "We run your unreleased checkpoint or agent on our jobs and send a failure report."
- What it does
-
Official documentation · 20 Oct 2026
Home page: "We run your unreleased checkpoint or agent on our jobs and send a failure report."
- Sold within
-
Private evals are requested through the home page; terms are scoped per engagement.
- Name
-
Official documentation · 20 Oct 2026
Name as published on the product page.
- URL
-
Official documentation · 20 Oct 2026
Product page, read 2026-10-20.
- Deployment
-
Official documentation · 20 Oct 2026
The pages read state no deployment model beyond the vendor-run service; saas is recorded as the vendor-operated reading.
Open TensorTest in the directory
Something wrong here? Send a correction — quote this product id: tensortest-private-evals.