NexAtlas Quick Verdict.
A fast read before you dig in — based on public documentation, not hands-on testing.
Best for
Platform teams with heavy GitHub Actions usage who want autoscaling runners without maintaining the fleet.
Not ideal for
Projects running on GitLab CI or Jenkins, or teams whose CI volume is too low to notice the cost.
Key strengths
- Autoscaling removes idle-runner waste
- Bare-metal runners speed up CPU-heavy jobs
- Drop-in: keep GitHub Actions, change one label
Things to consider
- Public pricing detail is thin; you need a quote
- GitHub Actions only — no GitLab or Jenkins orchestration
NexAtlas has not hands-on tested Tenki Cloud yet. This summary condenses public documentation and vendor materials checked on Oct 22, 2026; scores and first-hand user reviews will be added once independent testing is done.
What is Tenki Cloud.
Tenki Cloud sells compute for GitHub Actions. Instead of paying GitHub's per-minute rate for hosted runners, or hand-managing your own runner fleet, you install Tenki's GitHub App and point jobs at its autoscaling runner groups, which spin up on bare metal when work arrives and scale to zero when it does not.
The pitch is arithmetic: CI bills grow with test volume, and most teams run the same few heavyweight jobs over and over. Tenki advertises up to 30% faster workflows and up to 90% lower cost than GitHub-hosted runners, plus the control that comes with self-hosted infrastructure — custom images, bigger machines and network placement.
Key features.
The parts of the product that do the work
Autoscaling runner groups
Runners appear when a job is queued and disappear afterwards. You stop paying for idle machines and stop queueing behind concurrency limits.
Bare-metal performance
Jobs run on dedicated hardware rather than shared virtual runners, so CPU-heavy builds, test shards and Docker layers behave predictably.
Drop-in for existing workflows
You keep GitHub Actions as the orchestrator and change the runs-on target. No workflow rewrite, no migration of your pipeline definition.
Cost visibility and control
Usage-based billing with runner size options, so expensive jobs can be matched to the hardware they actually need instead of one default size.
How it works.
From input to finished output
Install the GitHub App
Authorise Tenki against your organisation or repositories and choose the runner groups to expose.
Point workflows at Tenki
Change runs-on to the Tenki runner label. Existing matrix builds, caches and secrets keep working.
Tune size and concurrency
Match CPU and memory to each job's bottleneck — a bigger runner is not automatically faster if the job is single-threaded.
Who it's for.
Typical teams and workflows
Own CI cost and reliability, and want autoscaling without maintaining a Kubernetes-based runner controller.
Shard tests across many parallel runners to cut wall-clock time on every pull request.
Free or low-cost capacity for public repositories where minute bills otherwise get donated away.
Pricing.
Public pricing checked on Oct 22, 2026
Tenki runs a freemium, usage-based model: start small on free capacity, then pay for runner minutes as volume grows. Exact per-minute rates are quoted through the site rather than published in full.
Pricing not verified. Tenki prices runner minutes by quote. NexAtlas has not verified current rates, so this page does not publish per-minute prices.
| Plan | Price | What you get |
|---|---|---|
| Freemium | $0 to start | Free allowance for small projects and evaluation, listed as freemium on directory sites. |
| Usage-based | Per runner minute | Billed on compute consumed. Machine sizes and concurrency determine the effective cost per pipeline. |
Swipe to see the full table.
Cost comparisons with GitHub-hosted runners depend on your job mix. Model your three most expensive workflows — minutes, cores and cache behaviour — before switching, and ask about egress, storage and support terms.
Pros & cons.
Strengths
- Autoscaling removes idle-runner waste
- Bare-metal runners speed up CPU-heavy jobs
- Drop-in: keep GitHub Actions, change one label
- Self-hosted control over images and network
Trade-offs
- Public pricing detail is thin; you need a quote
- GitHub Actions only — no GitLab or Jenkins orchestration
- Third-party runner fleet is another vendor in the trust chain
- Cost claims are vendor-published and workload-dependent
Alternatives worth comparing.
Other tools in this Top 10
- #06 · AutomationKlarentDescribe a test in plain English; Klarent's agents build, run and repair it across web, iOS and Android.
- #09 · CodingEmergent.shDescribe an app in plain language and Emergent builds, hosts and deploys it — database APIs included, no local setup.
- #07 · MarketingJason AIA hired AI SDR from Reply.io: multichannel outbound that writes, sequences and follows up on its own, then books the meeting.
FAQ.
Quick answers before you sign up
Do I have to rewrite my workflows?
No. GitHub Actions stays as the orchestrator; you change the runner label that jobs target. Your YAML structure, secrets and caches remain.
Is it cheaper than GitHub-hosted runners?
For high-volume CI, Tenki claims up to 90% lower cost and 30% faster runs. Real savings depend on your job mix and runner sizes, so measure rather than assume.
Does it work for public repositories?
Yes — that is one of the common use cases, since open-source CI minutes are otherwise a recurring donation. Check the current free allowance for public repos.
Sources & notes.
- Tenki Cloud — official website — tenki.cloud/
- Tenki Cloud — GitHub Actions runners guide — tenki.cloud/blog/github-actions-runners-beginners-guide
- Tenki Cloud — choosing a runner — www.tenki.cloud/docs/runner-offering
- FutureTools — Tenki Cloud listing — futuretools.io/tools/tenki-cloud
- AI Tool Navio — Tenki Cloud overview — aitoolnavio.com/tool/coding-development/devops-cloud/tenki-cloud
This page is an editorial summary compiled from the sources below — including Bocha (博查) web-search results and the tool’s own public pages — checked on Oct 22, 2026. Pricing, features and limits change frequently; confirm them on the official website before you buy. Atlas is an independent prototype and does not receive payment for this listing.