Skip to main content
Inferix
Pricing

Leveling up AI building and compute

Start for free, upgrade for higher quotas and premium storage, rent real GPUs by the second, and move to Enterprise for governance and compliance controls.

Explore the platform

Free

$0
Get started
  • 10GB private storage
  • Unlimited public repos
  • Community support
  • Public Spaces
Most popular

For individual builders

Pro

$9.99/mo
Subscribe to Pro
  • 100GB private storage
  • 100 GPU hours / month
  • 8× Spaces quota + priority queue
  • Private dataset viewer

For growing organizations

Team

$20/mo/user
Subscribe to Team
  • 500GB private storage / user
  • Unlimited private repos
  • Centralized billing
  • Organization-wide access controls
  • Priority support response queue

For mission-critical workloads

Enterprise

Custom
Talk to sales
  • SSO (OIDC) and enterprise RBAC
  • Audit logs and compliance exports
  • GPU data-residency enforcement
  • Dedicated support and onboarding

Need team governance, SSO, and audit controls? Talk to our Enterprise team →

Storage

Transparent, volume-based pricing

Store models, datasets, and artifacts at a flat, honest per-TB rate — with automatic volume discounts as you grow. No egress surprises, no per-request line items.

AWS S3
$23/TB
Backblaze Overdrive
$15/TB
Inferix
$12/TB

Comparative per-TB/mo list prices are indicative. Inferix figure is its real base storage rate.

Base

Up to 50 TB

Private$30/TB
Public$12/TB

Scale

−20%

50 TB+

Private$24/TB
Public$9.60/TB

Growth

−25%

200 TB+

Private$22.50/TB
Public$9/TB

Enterprise

−33%

500 TB+

Private$20.10/TB
Public$8.04/TB

Per-TB rates shown monthly. Base private rate ($0.03/GB) is live; volume-tier discounts and public rates are indicative.

Storage pricing calculator

Estimate what you'll pay in storage overage beyond your plan's included quota. Overage is billed at $0.03/GB per month.

Storage used150 GB
0 GB2,000 GB

Included with Pro

100 GB

Overage this month

$1.50

Need team governance, SSO, and audit controls?

Move to Team or Enterprise when you need organization-wide policy controls and compliance workflows.

Explore Enterprise
On-demand compute

Rent GPUs by the second. No commitments.

Spin up a real GPU in seconds, pay only for the time you use, and shut it down whenever you like.

GPUVRAMPrice / hr
RTX 5090Live rate
32 GB$0.64 / GPU-hrRent now
RTX 4090
24 GB$0.44 / GPU-hrRent now
L40S
48 GB$1.10 / GPU-hrRent now
RTX PRO 6000 Blackwell
96 GB$1.79 / GPU-hrRent now
A100
80 GB$1.89 / GPU-hrRent now
H100
80 GB$2.49 / GPU-hrRent now
H200
141 GB$2.99 / GPU-hrRent now
Per-second billing. You are only charged while your instance is running.Indicative rates — final pricing may vary. The RTX 5090 rate is live from the marketplace.
Spaces Hardware

On-demand hardware for your Spaces. Starting at $0.

Spaces are the easiest way to host and share ML apps and demos — build with Gradio, Streamlit, Docker, or static, attach on-demand hardware, and share with a single URL.

NameCPUMemoryAcceleratorVRAMHourly
CPU Basic2 vCPU16 GB——FREE
CPU Upgrade8 vCPU32 GB——$0.03
ZeroGPUdynamicdynamicNvidia RTX Pro 6000 Blackwellup to 96 GBFREE
Nvidia T4 - small4 vCPU15 GBNvidia T416 GB$0.40
Nvidia T4 - medium8 vCPU30 GBNvidia T416 GB$0.60
1x Nvidia L48 vCPU30 GBNvidia L424 GB$0.80
4x Nvidia L448 vCPU186 GBNvidia L496 GB$3.80
1x Nvidia L40S8 vCPU62 GBNvidia L40S48 GB$1.80
4x Nvidia L40S48 vCPU382 GBNvidia L40S192 GB$8.30
8x Nvidia L40S192 vCPU1534 GBNvidia L40S384 GB$23.50
Nvidia A10G - small4 vCPU15 GBNvidia A10G24 GB$1.00
Nvidia A10G - large12 vCPU46 GBNvidia A10G24 GB$1.50
2x Nvidia A10G - large24 vCPU92 GBNvidia A10G48 GB$3.00
4x Nvidia A10G - large48 vCPU184 GBNvidia A10G96 GB$5.00
Nvidia A100 - large12 vCPU142 GBNvidia A10080 GB$2.50
4x Nvidia A10048 vCPU568 GBNvidia A100320 GB$10.00
8x Nvidia A10096 vCPU1136 GBNvidia A100640 GB$20.00
Customon demandon demandon demandon demandon demand
Launch a SpaceIndicative rates — final pricing may vary, and availability varies by tier. The marketplace shows what is live right now; anything not listed there is not yet provisionable. FREE tiers are subject to fair-use quotas.
Inference Endpoints

Production endpoints, billed by the GPU.

Deploy any model to a dedicated, auto-scaling HTTPS endpoint. You pay the same hourly GPU rate as above — and endpoints can scale to zero, so an idle endpoint costs nothing.

Starter

from$0.44/hr

RTX 4090 · 24 GB

Small models, embeddings, and prototypes. Scales to zero when idle so you only pay while serving.

Featured

Standard

from$1.10/hr

L40S / A100 · 48–80 GB

7B–70B models in production with auto-scaling replicas and a stable HTTPS endpoint.

Frontier

from$2.49/hr

H100 / H200 · 80–141 GB

The largest open models and highest throughput. Multi-GPU capacity on demand.

Usage-based: an endpoint's cost is the hourly rate of the GPU it runs on, prorated per second. Non-5090 GPU rates are indicative drafts.

Deploy an endpoint
Inference Endpoints

Dedicated inference, starting at $0.033/hour.

Deploy any model on dedicated, autoscaling infrastructure — production-ready, with no cold starts.

Learn more
CPU instances
InstancevCPUMemoryHourly
Intel Sapphire Rapids (aws)12GB$0.03
Intel Sapphire Rapids (aws)24GB$0.07
Intel Sapphire Rapids (aws)48GB$0.13
Intel Sapphire Rapids (aws)816GB$0.27
Intel Sapphire Rapids (aws)1632GB$0.54
Intel Xeon (azure)12GB$0.06
Intel Xeon (azure)24GB$0.12
Intel Xeon (azure)48GB$0.24
Intel Xeon (azure)816GB$0.48
Intel Sapphire Rapids (gcp)12GB$0.05
Intel Sapphire Rapids (gcp)24GB$0.10
Intel Sapphire Rapids (gcp)48GB$0.20
Intel Sapphire Rapids (gcp)816GB$0.40
Accelerators instances
InstanceSizeMemoryHourly
Inf2 Neuron (aws)x114.5GB$0.75
Inf2 Neuron (aws)x1-large124GB$1.95
Inf2 Neuron (aws)x12760GB$12.00
TPU v5e (gcp)1x116GB$1.20
TPU v5e (gcp)2x264GB$4.75
TPU v5e (gcp)2x4128GB$9.50
GPU instances
InstanceCountMemoryHourly
NVIDIA T4 (aws)114GB$0.50
NVIDIA T4 (aws)456GB$3.00
NVIDIA L4 (aws)124GB$0.80
NVIDIA L4 (aws)496GB$3.80
NVIDIA L40S (aws)148GB$1.80
NVIDIA L40S (aws)4192GB$8.30
NVIDIA L40S (aws)8384GB$23.50
NVIDIA A100 (aws)180GB$2.50
NVIDIA A100 (aws)2160GB$5.00
NVIDIA A100 (aws)4320GB$10.00
NVIDIA A100 (aws)8640GB$20.00
NVIDIA H100 (aws)180GB$4.50
NVIDIA H100 (aws)2160GB$9.00
NVIDIA H100 (aws)4320GB$18.00
NVIDIA H100 (aws)8640GB$36.00
NVIDIA H200 (aws)1141GB$5.00
NVIDIA H200 (aws)2282GB$10.00
NVIDIA H200 (aws)4564GB$20.00
NVIDIA H200 (aws)81128GB$40.00
NVIDIA B200 (aws)1179GB$9.25
NVIDIA B200 (aws)2358GB$18.50
NVIDIA B200 (aws)4716GB$37.00
NVIDIA B200 (aws)81432GB$74.00
NVIDIA RTX PRO 6000 (aws)196GB$2.75
NVIDIA RTX PRO 6000 (aws)2192GB$5.50
NVIDIA RTX PRO 6000 (aws)4384GB$11.00
NVIDIA RTX PRO 6000 (aws)8768GB$22.00

Indicative rates — final pricing may vary.

Volume & committed use

Reserve capacity or commit to volume for up to ~40% off on-demand rates. Discount is indicative — let's find the right fit for your workload.

Talk to us
Compare plans

Compare plans in detail

Every feature, side by side.

FeatureFreeProTeamEnterprise
Private storage10GB100GB500GB / userCustom
GPU inference hoursPay as you go100 / month500 / monthCustom
Spaces quotaStandard8× priority8× priorityCustom
Private dataset viewer
Organization-wide access controls
SSO (OIDC)
Audit logs
Dedicated support
Inferix Pro

Do more with Inferix Pro

Upgrade your Inferix experience for $9.99/month.

Private storage

10× private storage capacity

Public storage

2× public storage capacity

Inference Providers

20× included inference credits

Spaces

Host your own ZeroGPU, Gradio & Docker Spaces on compute

ZeroGPU

8× ZeroGPU quota and highest queue priority

Spaces Dev Mode

Fast iterations via SSH/VS Code

Blog

Personal blog publishing

Dataset Viewer

For private datasets

Features Preview

Get early access to upcoming features

PRO Badge

Show your support with a PRO badge

Get Pro
InferixPROInferixPROInferixPROInferixPROInferixPROInferixPROInferixPROInferixPROInferixPROInferixPRO
The Hub

Explore, experiment, collaborate and build with Machine Learning

The Inferix Hub is the central place to explore, experiment, collaborate and build with real GPUs — join the open ML movement.

Sign up
inferix.co

sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2

Sentence Similarity

↓ 52.1M

amazon/chronos-2

Time-Series

↓ 21.4M

facebook/opt-125m

Text Generation

↓ 17.1M

Rixy Playground

Gradio · ZeroGPU

Image Studio

Docker · L40S

Inferix brand assets

Logos, colors, and the Rixy render-cube mascot — everything you need to build with the Inferix brand.

Brand assetsEnter the Inferix Realm

    We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy