Segmind review
Updated September 28, 2026 · DeployCue Editorial Team · Facts checked September 26, 2026
We earn a commission when you sign up through links on this site. Partner status can affect where a brand appears in our rankings. Partners with a live offer in your country come first, then every other brand by score.
Our verdict
Segmind suits small teams and developers who want metered GPU and serverless access, with SOC 2 certification on file. The per second billing model is the separator: you pay for what a job actually runs, and GPU hour from $0.68 sits near the €0.63 median. Free credits are no, so expect a paid start.
Key facts
- Best for
- Small teams wanting metered, per second GPU and serverless access with SOC 2 assurance.
- GPU models
- NVIDIA H100, NVIDIA A100, NVIDIA L40, NVIDIA A40, NVIDIA A5000, NVIDIA A6000↗
- GPU hour from
- $0.68↗
- Billing granularity
- Per second↗
- Free credits
- No free signup credits; paid plans include monthly credits: Pro $50, Business $99, Scale $599↗
- Interconnect
- No dedicated interconnect product; multi-cloud GPU endpoints on AWS, Lambda, GCP and Azure↗
- Compliance certifications
- SOC 2↗
- Legal entity
- Segmind Inc.↗
- Headquarters
- United States↗
- Support channels
- Email, Community, Help-centre↗
- Pricing model
- Hourly billing
- Official website
- segmind.com
Facts checked September 26, 2026, 5 sources
Pros and cons
Pros
- GPU hour from $0.68 compares well with the €0.61 median for GPU providers
- Monthly credits are bundled into paid plans, so part of your usage is prepaid
- Carries SOC 2 certification, which matters in vendor reviews
Cons
- Free credits are no, so sign-up brings nothing to try and a paid plan supplies the monthly credits
- Interconnect is delivered as multi-cloud GPU endpoints on AWS, Lambda, GCP and Azure, which suits teams already anchored to one of those clouds
- Support runs on email, community and help centre only, with no live channel listed
Who it suits
- Developers running short inference or fine-tuning jobs on named NVIDIA cards
- Small teams that want metered serverless access instead of a reserved commitment
- Readers who place weight on SOC 2 certification in a vendor review
Who should look elsewhere
- Readers after parts outside the NVIDIA H100, NVIDIA A100 and NVIDIA L40 listed here, where ComputePrices.com carries a much longer catalogue
- Teams needing cards such as B300 or V100, which SberCloud lists outside the NVIDIA H100, NVIDIA A100 and NVIDIA L40 on offer
- Buyers who want credits on sign-up, since free credits are no on the plans here and Salad runs a different community engine
What it does well
The catalogue covers NVIDIA H100, NVIDIA A100 and NVIDIA L40, so you can move between current data centre cards and older parts without changing provider. Billing is per second, which suits bursty inference and short training runs. On network reach, interconnect comes as multi-cloud GPU endpoints on AWS, Lambda, GCP and Azure, so the service fits teams already working inside one of those clouds.
Who runs it
Segmind Inc. is the legal entity behind the platform, and the headquarters country is listed as the US, which is what a procurement checklist will ask for. Support runs on email, community and help centre: help centre articles first, then email or the community forum. Written answers suit account and billing questions, though a production incident may need patience.
How Segmind compares
Segmind vs RunPod8.9 vs 9.5
Segmind vs ScalaHosting8.9 vs 8.3
Segmind vs Contabo8.9 vs 9.2
Segmind vs Kamatera8.9 vs 9.5
Segmind vs Cloudways8.9 vs 9.2
Segmind vs PlanetHoster8.9 vs 8.7
Segmind vs Acronis8.9 vs 9.9
Segmind vs Aruba8.9 vs 9.3
Segmind vs Decodo8.9 vs 9.1
Segmind vs Hostinger8.9 vs 9.1
Segmind vs MarsProxies8.9 vs 9.0
Segmind vs hide.me8.9 vs 8.7
Segmind vs Proton Drive8.9 vs 8.3
Frequently asked questions
What does a GPU hour cost on Segmind?+
A GPU hour starts at $0.68, and that sits close to the €0.63 median across the GPU providers listed here. Short bursts therefore cost about what the field charges rather than a hyperscaler premium.
How finely are GPU jobs billed?+
Billing is per second, one of the finer cutoffs in the set, and only 26.4% of listed providers meter that tightly. A short job no longer pays for a whole block, though idle time keeps accruing.
Is there a free tier or trial credit?+
There are no free signup credits, so evaluation starts on a paid plan. Each paid tier ships monthly credits, with the Pro tier carrying the smallest bundle and Scale the largest, and usage is metered hourly, which softens early spend while keeping the trial billable.
Which GPUs can I rent?+
The catalogue runs to NVIDIA H100, NVIDIA A100 and NVIDIA L40, mixing current data centre cards with older parts on the same account. Readers who need the newest silicon or consumer-grade cards will find longer lists elsewhere, including different hardware classes.
Who runs Segmind and what stands behind it?+
Segmind operates as Segmind Inc. and lists US as its headquarters country, alongside SOC 2 certification. Both matter most when a client or auditor asks for a named contracting company and a completed security review before workloads move over.
How do I reach support?+
Support runs on email, community and help centre: help centre articles, email and a community forum. The help centre is the fastest route for setup and model questions, while email suits anything tied to your account, invoices or credits. There is no live channel listed, so plan for written replies.
Can I connect my own cloud directly?+
Interconnect is delivered as multi-cloud GPU endpoints on AWS, Lambda, GCP and Azure, so you can place NVIDIA H100, NVIDIA A100 and NVIDIA L40 work next to data you already hold in those clouds. That suits teams committed to a single provider, and placement and network terms are worth confirming with the team before you commit.
Sources
- docs.segmind.com/docs/get-started/key-concepts checked September 26, 2026
Backs: Billing granularity, Interconnect
“For billing, Segmind measures the exact GPU processing time used to execute your request”
- segmind.com/pricing checked September 26, 2026
Backs: Compliance certifications, Free credits, GPU hour from, GPU models, Pricing model
“SOC 2 compliance”
- segmind.com/privacy checked September 26, 2026
Backs: Headquarters
“Segmind Inc. Rohit Ramesh 1013, Center Road, Suite 403-B Wilmington, DE 19805 United States”
- segmind.com/support checked September 26, 2026
Backs: Support channels
“Join our Discord community for platform questions, API help, and troubleshooting ... email the support team ... Browse the docs”
- segmind.com/terms checked September 26, 2026
Backs: Legal entity
“This Main Cloud Services Agreement is entered into ... between Segmind Inc.”
Compiled from 5 source pages, checked September 26, 2026. We did not open an account.
Not confirmed yet: Egress fee.
Related by the facts
Segmind
Small teams wanting metered, per second GPU and serverless access with SOC 2 assurance.