Hyperbolic · Pricing Plans

Hyperbolic Ai Plans Pricing

Reconciled plans and per-token / per-image / per-GPU-hour pricing for the Hyperbolic Serverless Inference and GPU Marketplace services. Source: hyperbolic.ai, docs.hyperbolic.ai, and third-party pricing trackers as of May 2026.

Hyperbolic Ai Plans Pricing is the machine-readable pricing-plan profile for Hyperbolic on the APIs.io network, conforming to the API Commons Plans specification.

It defines 16 plans, covering tier, usage-based, commitment, and commission tiers, with named plans including Basic, Pro, Enterprise, Llama 3.1 405B Inference, DeepSeek R1 Inference, and 11 more.

Tagged areas include AI, GPU, Inference, and Marketplace.

16 Plans API Commons Plans
View Source
AIGPUInferenceMarketplace

Plans

Basic tier

Free tier — 60 requests per minute, community support, no minimum deposit.

Request rate (requests · minute) 0.00 USD
Pro tier

Pro tier — 600 requests per minute, priority and email support, requires at least $5 deposit.

Request rate (requests · minute) 0.00 USD
Minimum deposit (usd · usage) 5.00 USD
Enterprise tier

Enterprise tier — unlimited rate limits, 24/7 dedicated support, custom contracts.

Request rate (requests · minute) Call USD
Llama 3.1 405B Inference usage-based

Per-million-token pricing for Meta-Llama-3.1-405B-Instruct.

Tokens (input + output) (tokens · usage) 4.00 USD
DeepSeek R1 Inference usage-based

Per-million-token pricing for DeepSeek-R1.

Tokens (input + output) (tokens · usage) 3.00 USD
Llama 3.3 70B Inference usage-based

Per-million-token pricing for Llama-3.3-70B and similar 70B-class models.

Tokens (input + output) (tokens · usage) 0.40 USD
Small LLMs Inference usage-based

Per-million-token pricing floor for 7B–8B class models.

Tokens (input + output) (tokens · usage) 0.10 USD
Vision Models Inference usage-based

Per-million-token pricing floor for multimodal vision models (Llama 3.2 Vision, Qwen2-VL).

Tokens (input + output) (tokens · usage) 0.15 USD
Image Generation usage-based

Per-image pricing floor for SDXL / SD3.5 / FLUX diffusion models.

Images generated (image · usage) 0.0025 USD
Audio TTS usage-based

Per-1000-character pricing floor for text-to-speech.

Characters synthesized (character · usage) 0.001 USD
GPU On-Demand — RTX 4090 usage-based

Per-GPU-hour pricing for RTX 4090 on-demand rental.

GPU hours (gpu_hour · hour) 0.50 USD
GPU On-Demand — A100 usage-based

Per-GPU-hour pricing for NVIDIA A100.

GPU hours (gpu_hour · hour) 1.80 USD
GPU On-Demand — H100 usage-based

Per-GPU-hour pricing for NVIDIA H100. Marketplace pricing ranges from $1.39 to $3.20 per hour depending on supplier, configuration, and dedicated vs shared.

GPU hours (entry marketplace) (gpu_hour · hour) 1.39 USD
GPU hours (dedicated) (gpu_hour · hour) 3.20 USD
GPU On-Demand — H200 usage-based

Per-GPU-hour pricing for NVIDIA H200 on-demand rental.

GPU hours (gpu_hour · hour) Marketplace USD
Reserved Clusters commitment

3-12 month prepaid GPU cluster commitments with volume discounts up to 40% off marketplace on-demand pricing.

Volume discount (percent · usage) -40 USD
Marketplace Platform Fee commission

Hyperbolic charges a 10% platform fee on rental income from GPU suppliers.

Platform fee (percent · usage) 10 USD

Sources