All Hyperbolic Plans & Pricing

Plan Monthly Annual Best For
View all features by plan (compare side-by-side)

RTX 3070

  • RTX 3070 GPU
  • Hourly billing, no minimum commitment
  • On-demand availability
  • No long-term contracts
  • Starting at $0.16/hr

RTX 4090

  • RTX 4090 24GB VRAM
  • Hourly billing at $0.30/hr
  • On-demand availability
  • No minimum commitment
  • OpenAI-compatible serverless inference also available

H100 SXM

  • H100 SXM 80GB VRAM
  • Hourly billing at $1.50/hr
  • High-bandwidth NVLink interconnect
  • On-demand availability
  • No minimum commitment

H200

  • H200 GPU
  • Hourly billing at $2.40/hr
  • On-demand availability
  • High-bandwidth interconnect
  • No minimum commitment

B200

  • B200 GPU
  • Hourly billing at $3.50/hr
  • On-demand availability
  • Top-tier compute performance
  • No minimum commitment
Compare Hyperbolic with alternativesAdjust seats, lock a tier, add up to 2 more products side-by-side. Shareable URL.
Quick Answer
Last verified:
Medium confidence

Hyperbolic costs $0.16 to $3.50 per GPU/hour as of July 2026, with 5 plans available. Pricing depends on your chosen tier, contract length, and negotiated discounts.

Use the interactive pricing calculator to estimate your exact cost based on team size and requirements.

  • Free tier: No free tier available

Hyperbolic offers 5 pricing tiers: RTX 3070, RTX 4090, H100 SXM, H200, B200. The RTX 4090 plan is cost-sensitive inference and smaller model fine-tuning.

Compared to other ai/gpu cloud compute software, Hyperbolic is positioned at the budget-friendly price point.

  • 6 documented hidden costs beyond list price

How much does Hyperbolic cost?

Hyperbolic pricing ranges from $0.16 to $3.50/GPU/hour across 5 plans. Plans include RTX 3070 (custom pricing), RTX 4090 (custom pricing), H100 SXM (custom pricing), H200 (custom pricing), B200 (custom pricing).

Hyperbolic Pricing Overview

Hyperbolic has 5 pricing plans ranging from $0.16 to $3.50/GPU/hour. The RTX 3070 plan requires contacting sales for a custom quote and is designed for budget inference and lightweight fine-tuning workloads. The RTX 4090 plan requires contacting sales for a custom quote and is designed for cost-sensitive inference and smaller model fine-tuning. The H100 SXM plan requires contacting sales for a custom quote and is designed for large-scale training runs and high-throughput batch inference. The H200 plan requires contacting sales for a custom quote and is designed for frontier model training and high-memory inference workloads. The B200 plan requires contacting sales for a custom quote and is designed for most demanding ai training and inference at the highest performance tier.

There are at least 6 documented hidden costs beyond Hyperbolic's list price, including implementation, training, and add-on fees.

This pricing was last verified in June 9, 2026 from 1 independent source.

Hyperbolic is an on-demand GPU cloud marketplace where compute is billed by the hour. Prices range from $0.16/hr for entry-level GPUs up to $3.50/hr for B200 instances as of 2026, with no contracts or minimum commitments.

The platform offers a range of GPUs — from consumer cards like the RTX 4090 at $0.30/hr to data center hardware including the H100 SXM at $1.50/hr, H200 at $2.40/hr, and B200 at $3.50/hr. All tiers are billed hourly based on actual usage.

In addition to dedicated GPU rentals, Hyperbolic provides a serverless inference API that is OpenAI-compatible, letting developers call popular open-source models (Llama, DeepSeek, Qwen) without managing GPU infrastructure directly.

How Hyperbolic Pricing Compares

Compare Hyperbolic pricing against top alternatives in AI/GPU Cloud Compute.

Compare Hyperbolic vs Alternatives

Before committing to Hyperbolic, compare pricing with these 3 alternatives in the same category.

All Hyperbolic alternatives & migration guides

What Companies Actually Pay for Hyperbolic

Median per-1M-token pricing across 3 models
Input $1.25/1M
Output $1.25/1M
Flagship models in this provider's catalog
Model Input /1M Output /1M Blended /1M
hyperbolic_qwen3-coder-480b-a35b-instruct_fp8 $2.00 $2.00 $2.00
hyperbolic_deepseek-v3-0324 $1.25 $1.25 $1.25
hyperbolic_llama-3-3-instruct-70b $0.400 $0.400 $0.400
Review scores
Trustpilot 4.5/5 (116)
Top pricing complaints
One user reported minimal improvements after following the program for six months and receiving no reply from the company for support.Hyperbolic Global is widely reported as an alleged scam
Source: Artificial Analysis — medians aggregated from 3 models in this provider's catalog. Per-1M-token pricing reflects list rates.

How Hyperbolic Pricing Compares

Software Starting Price Top Price
Hyperbolic $0.16/GPU/hour $3.5/GPU/hour
CoreWeave $6.27/instance/hour $68.8/instance/hour
Lambda $0.69/GPU/hour $6.99/GPU/hour
Modal Free $250/GPU/hour
Paperspace $0.45/GPU/hour $5.95/GPU/hour
RunPod $0.27/GPU/hour $7.39/GPU/hour

6 Hyperbolic Hidden Costs Beyond the List Price

Beyond the listed price, Hyperbolic has at least 6 documented hidden costs that can significantly increase total cost of ownership.

Watch for 6 hidden costs
  • GPU Idle Waste 20% to 40%
    high 1 source
    industry "This idle time still incurs charges, accumulating to thousands of dollars monthly for multi-node clusters"
  • Data Egress and Transfer Fees 50% to 100%
    critical 1 source
    industry "For data-heavy deployments, egress can increase invoices by 50% to 100%"
  • Storage Overages $2,000–$2,300 per year
    high 1 source
    industry "Storing 10TB of training data in cloud services can cost approximately $2,000–$2,300 per year, excluding transfer or access fees"
  • API and Managed Service Fees
    medium 1 source
    industry "API and Managed Service Fees: Per-use fees for orchestration, monitoring tools, and managed services can add unexpected layers of expense"
  • Networking and Interconnect Charges $0.01/GB
    medium 1 source
    industry "For instance, AWS charges $0.01/GB for cross-AZ data transfer in each direction"
  • Software and Licensing
    medium 1 source
    industry "API and Managed Service Fees: Per-use fees for orchestration, monitoring tools, and managed services can add unexpected layers of expense"
Tip

Ask your Hyperbolic sales rep about these costs upfront. Getting them in writing before signing can save you from surprise charges later.

Full hidden costs breakdown →

Intelligence sourced from 2 independent sources
industry Trustpilot Consumer reviews
Key claims include inline source attribution. Data verified against multiple independent sources. 7 source citations total.

Hyperbolic Pricing FAQ

01 How much does Hyperbolic cost?

Hyperbolic charges by the hour based on which GPU you rent. Prices currently range from $0.16/hr for entry-level GPUs to $3.50/hr for B200 instances. There are no monthly subscriptions or minimum commitments.

02 How much does an H100 cost on Hyperbolic?

The H100 SXM is currently priced at $1.50/hr on Hyperbolic's marketplace. Pricing is refreshed periodically based on supplier rates, so check hyperbolic.ai/marketplace for the latest figures.

03 What is the cheapest GPU available on Hyperbolic?

Entry-level options start at $0.16/hr, making Hyperbolic accessible for developers who need affordable GPU access for inference or smaller workloads. The RTX 4090 is available at $0.30/hr.

04 Does Hyperbolic offer a free plan or free credits?

New users receive a small promotional credit on sign-up, but it is restricted to the serverless inference API and cannot be applied to GPU rentals. A minimum deposit is required to unlock dedicated GPU access.

05 How does Hyperbolic compare to other GPU clouds?

Hyperbolic's H100 SXM is listed at $1.50/hr, which is notably lower than older published rates from some competitors. For the most accurate comparison, check each provider's live marketplace since GPU rental rates shift frequently.

06 Does Hyperbolic support serverless inference?

Yes. Hyperbolic offers a pay-per-token serverless API for popular open-source models including Llama, DeepSeek, and Qwen. It uses an OpenAI-compatible endpoint, so teams can integrate it without changing existing tooling.

Is this pricing incorrect? — we'll verify and update it.