Home
Hyperbolic
Hyperbolic
Hyperbolic is an open-access AI cloud and decentralized GPU marketplace serving 200,000+ builders with affordable inference and bare-metal compute. The platform combines a serverless OpenAI-compatible inference API spanning 25+ open-source LLMs (including the only public Llama-3.1-405B-Base in BF16), image and audio models, with an on-demand GPU rental marketplace aggregating idle H100 / H200 / A100 / RTX 4090 capacity from third-party suppliers at 3-10x lower cost than hyperscalers. Reserved clusters, dedicated endpoints, an OpenAI-drop-in Python and TypeScript SDK, a Go CLI, an MCP server, the Hyperbolic AgentKit, the open-source Hyper-dOS distributed operating system, and Coinbase x402 crypto payments round out the stack.
6 APIs
5 Capabilities
21 Features
AI Artificial Intelligence Compute Decentralized DePIN GPU Image Generation Inference LLM Marketplace Open Source
Hyperbolic publishes 5 APIs on the APIs.io network, including Chat Completions API, Completions API, Image Generation API, and 2 more. Tagged areas include AI, Artificial Intelligence, Compute, Decentralized, and DePIN.
The Hyperbolic catalog on APIs.io includes 5 machine-runnable capabilities , 1 JSON-LD context, and 1 Spectral governance ruleset.
Hyperbolic’s developer surface includes developer portal, documentation, getting-started guide, signup flow, sandbox, engineering blog, support, and 40 more developer resources.
OpenAI-compatible chat completions endpoint serving 25+ open-source LLMs including Llama 3.1 8B/70B/405B, Qwen 2.5, DeepSeek V3, DeepSeek R1, Hermes 3, Mistral, and vision model...
Legacy OpenAI-compatible text completions endpoint for base-model prompting. Notably exposes Llama-3.1-405B-Base in both BF16 (high-throughput precision) and FP8 (low-latency) —...
Generate images from text prompts using diffusion models including Stable Diffusion XL, Stable Diffusion 3.5, FLUX.1 [schnell] / [dev] (sunset), and ControlNet variants. Support...
Convert text to natural-sounding speech using Melo TTS (sunset) and Whisper (coming soon). POST /v1/audio/generation accepts text and speed; returns base64-encoded audio. Pricin...
OpenAI-compatible model discovery endpoint. GET /v1/models returns the catalog of currently served inference models — chat, base completion, image, vision, and audio — so client...
Decentralized on-demand GPU compute marketplace renting idle H100, H200, A100, and RTX 4090 capacity from third-party suppliers. Pricing starts at $0.50/GPU/hr (RTX 4090), $1.39...
Run Capabilities with Naftiko — Deploy and orchestrate these API capabilities using Naftiko Fleet.
Run with Naftiko
Hyperbolic Audio Generation API — Audio. 1 operations. Lead operation: Hyperbolic Generate Audio. Self-contained Naftiko capability covering one Hyperbolic business surface.
Run with Naftiko
Hyperbolic Chat Completions API — Chat. 1 operations. Lead operation: Hyperbolic Create A Chat Completion. Self-contained Naftiko capability covering one Hyperbolic business sur...
Run with Naftiko
Hyperbolic Completions API — Completions. 1 operations. Lead operation: Hyperbolic Create A Completion. Self-contained Naftiko capability covering one Hyperbolic business surface.
Run with Naftiko
Hyperbolic Image Generation API — Image. 1 operations. Lead operation: Hyperbolic Generate An Image. Self-contained Naftiko capability covering one Hyperbolic business surface.
Run with Naftiko
Hyperbolic Models API — Models. 1 operations. Lead operation: Hyperbolic List Models. Self-contained Naftiko capability covering one Hyperbolic business surface.
Run with Naftiko
Run Capabilities with Naftiko — Deploy and orchestrate these API capabilities using Naftiko Fleet.
Run with Naftiko
OpenAI-compatible REST API at https://api.hyperbolic.xyz/v1 — drop-in replacement requiring only api_key and base_url changes
25+ open-source inference models — Llama 3.1 8B/70B/405B, Qwen 2.5, DeepSeek V3, DeepSeek R1, Hermes 3, Mistral, Llama 3.2 Vision, Qwen2-VL
Only public provider serving Llama-3.1-405B-Base in BF16 (high-throughput precision) and FP8 (low-latency)
Image generation with Stable Diffusion XL, SD 3.5, FLUX.1 (sunset), ControlNet, custom LoRAs from $0.0025/image
Audio generation (Melo TTS sunset, Whisper coming soon) from $0.001 per 1000 characters
Decentralized GPU marketplace renting H100, H200, A100, RTX 4090 from third-party suppliers
On-demand instances deploy in under 1 minute — up to 75% cheaper than AWS / Azure / GCP
Reserved clusters with volume discounts up to 40% on 3-12 month commitments
Dedicated single-tenant endpoints for high-throughput inference workloads
Streaming, tool/function calling, structured JSON output, vision input on supported models
Python SDK and TypeScript SDK fully OpenAI-compatible
Hyperbolic CLI (Go) distributed via Homebrew tap
Model Context Protocol (MCP) server for Claude integration
Hyperbolic AgentKit — Python agent framework
Hyper-dOS — open-source Distributed Operating System for GPU orchestration
Coinbase x402 payment integration for crypto-native chat completions
Crypto payment support (USDC, USDT, DAI) alongside credit cards
Hugging Face Inference Providers integration
Three usage tiers — Basic (60 RPM, free), Pro (600 RPM, $5+ deposit), Enterprise (unlimited, sales)
200,000+ builder community
Marketplace platform fee of 10% on GPU rental income
0 classes · 8 properties
JSON-LD
9 rules ·
2 errors
7 warnings
SPECTRAL
Sources
aid: hyperbolic-ai
url: https://raw.githubusercontent.com/api-evangelist/hyperbolic-ai/refs/heads/main/apis.yml
apis:
- aid: hyperbolic-ai:hyperbolic-chat-completions-api
name: Hyperbolic Chat Completions API
tags:
- AI
- Chat
- Completions
- Inference
- LLM
humanURL: https://docs.hyperbolic.ai/inference/overview
properties:
- url: https://docs.hyperbolic.ai/inference/overview
type: Documentation
- url: https://docs.hyperbolic.ai/inference/chat-completion
type: Documentation
- url: openapi/hyperbolic-chat-completions-api-openapi.yml
type: OpenAPI
- url: json-schema/hyperbolic-chat-completion-schema.json
type: JSONSchema
- url: json-ld/hyperbolic-ai-context.jsonld
type: JSONLD
- type: NaftikoCapability
url: capabilities/inference-chat-completions.yaml
description: OpenAI-compatible chat completions endpoint serving 25+ open-source LLMs including Llama 3.1 8B/70B/405B,
Qwen 2.5, DeepSeek V3, DeepSeek R1, Hermes 3, Mistral, and vision models (Llama 3.2 Vision, Qwen2-VL).
Supports streaming, tool/function calling, structured JSON output, and chain-of-thought reasoning. Drop-in
OpenAI replacement — change api_key and base_url to https://api.hyperbolic.xyz/v1.
- aid: hyperbolic-ai:hyperbolic-completions-api
name: Hyperbolic Completions API
tags:
- AI
- Completions
- Inference
- LLM
humanURL: https://docs.hyperbolic.ai/inference/overview
properties:
- url: https://docs.hyperbolic.ai/inference/overview
type: Documentation
- url: openapi/hyperbolic-completions-api-openapi.yml
type: OpenAPI
- type: NaftikoCapability
url: capabilities/inference-completions.yaml
description: Legacy OpenAI-compatible text completions endpoint for base-model prompting. Notably exposes
Llama-3.1-405B-Base in both BF16 (high-throughput precision) and FP8 (low-latency) — Hyperbolic is the
only public provider serving the base model in BF16.
- aid: hyperbolic-ai:hyperbolic-image-generation-api
name: Hyperbolic Image Generation API
tags:
- AI
- Diffusion
- Image Generation
- Inference
humanURL: https://docs.hyperbolic.ai/inference/overview
properties:
- url: https://docs.hyperbolic.ai/inference/overview
type: Documentation
- url: openapi/hyperbolic-image-generation-api-openapi.yml
type: OpenAPI
- type: NaftikoCapability
url: capabilities/inference-image-generation.yaml
description: Generate images from text prompts using diffusion models including Stable Diffusion XL,
Stable Diffusion 3.5, FLUX.1 [schnell] / [dev] (sunset), and ControlNet variants. Supports custom LoRAs.
Pricing from $0.0025 per image. POST /v1/image/generation accepts model_name, prompt, steps, cfg_scale,
height, width.
- aid: hyperbolic-ai:hyperbolic-audio-generation-api
name: Hyperbolic Audio Generation API
tags:
- AI
- Audio
- Inference
- Text To Speech
humanURL: https://docs.hyperbolic.ai/inference/overview
properties:
- url: https://docs.hyperbolic.ai/inference/overview
type: Documentation
- url: openapi/hyperbolic-audio-generation-api-openapi.yml
type: OpenAPI
- type: NaftikoCapability
url: capabilities/inference-audio-generation.yaml
description: Convert text to natural-sounding speech using Melo TTS (sunset) and Whisper (coming soon).
POST /v1/audio/generation accepts text and speed; returns base64-encoded audio. Pricing from
$0.001 per 1000 characters.
- aid: hyperbolic-ai:hyperbolic-models-api
name: Hyperbolic Models API
tags:
- AI
- Inference
- Models
humanURL: https://docs.hyperbolic.ai/inference/overview
properties:
- url: https://docs.hyperbolic.ai/inference/overview
type: Documentation
- url: openapi/hyperbolic-models-api-openapi.yml
type: OpenAPI
- type: NaftikoCapability
url: capabilities/inference-models.yaml
description: OpenAI-compatible model discovery endpoint. GET /v1/models returns the catalog of currently
served inference models — chat, base completion, image, vision, and audio — so clients can route requests
by capability without hard-coded model IDs.
- aid: hyperbolic-ai:hyperbolic-marketplace-api
name: Hyperbolic GPU Marketplace API
tags:
- AI
- Compute
- GPU
- Infrastructure
- Marketplace
humanURL: https://docs.hyperbolic.ai/on-demand/overview
properties:
- url: https://docs.hyperbolic.ai/on-demand/overview
type: Documentation
- url: https://docs.hyperbolic.ai/on-demand/quickstart
type: Documentation
- type: GitHubRepository
url: https://github.com/HyperbolicLabs/hyperbolic-cli
- type: GitHubRepository
url: https://github.com/HyperbolicLabs/hyperbolic-mcp
description: Decentralized on-demand GPU compute marketplace renting idle H100, H200, A100, and RTX 4090
capacity from third-party suppliers. Pricing starts at $0.50/GPU/hr (RTX 4090), $1.39-$1.49/hr (H100),
up to $3.20/hr (H100 dedicated). Up to 75% cheaper than AWS/Azure/GCP. Instances deploy in under
one minute. Reserved clusters and dedicated single-tenant endpoints available. Hyperbolic charges a
10% platform fee on rental income. Managed via app.hyperbolic.ai dashboard, the `hyperbolic-cli`
Go CLI, the MCP server, and the Hyperbolic AgentKit.
name: Hyperbolic
tags:
- AI
- Artificial Intelligence
- Compute
- Decentralized
- DePIN
- GPU
- Image Generation
- Inference
- LLM
- Marketplace
- Open Source
kind: contract
image: https://kinlane-productions2.s3.amazonaws.com/apis-json/apis-json-logo.jpg
access: 3rd-Party
common:
- type: Portal
url: https://www.hyperbolic.ai/
- type: Documentation
url: https://docs.hyperbolic.ai/
- url: https://docs.hyperbolic.ai/overview
name: Platform Overview
type: Documentation
- url: https://docs.hyperbolic.ai/overview/platform-comparison
name: Platform Comparison
type: Documentation
- url: https://docs.hyperbolic.ai/inference/overview
name: Serverless Inference Overview
type: Documentation
- url: https://docs.hyperbolic.ai/inference/quickstart
name: Serverless Inference Quick Start
type: GettingStarted
- url: https://docs.hyperbolic.ai/on-demand/overview
name: On-Demand GPU Overview
type: Documentation
- url: https://docs.hyperbolic.ai/on-demand/quickstart
name: On-Demand GPU Quick Start
type: GettingStarted
- url: https://docs.hyperbolic.ai/reserved/overview
name: Reserved Clusters Overview
type: Documentation
- url: https://app.hyperbolic.ai/
name: Hyperbolic Console
type: SignUp
- url: https://app.hyperbolic.ai/settings/api-keys
name: API Keys
type: Documentation
- url: https://app.hyperbolic.ai/models
name: Inference Playground
type: Sandbox
- url: https://www.hyperbolic.ai/blog
type: Blog
- url: https://www.hyperbolic.ai/about
type: AboutUs
- url: https://www.hyperbolic.ai/careers
type: Careers
- url: https://jobs.ashbyhq.com/hyperbolic
name: Hyperbolic Jobs
type: Careers
- url: https://www.hyperbolic.ai/media-kit
name: Media Kit
type: PressKit
- url: https://www.hyperbolic.ai/privacy-policy
type: PrivacyPolicy
- url: https://www.hyperbolic.ai/terms-of-use
type: TermsOfService
- url: mailto:[email protected]
name: Contact
type: ContactForm
- url: mailto:[email protected]
name: Sales
type: ContactForm
- url: mailto:[email protected]
name: Support
type: Support
- url: https://calendly.com/d/cq79-jyv-jg4/hyperbolic-sales-demo
name: Sales Demo
type: ContactForm
- url: https://twitter.com/hyperbolic_labs
name: X / Twitter
type: Twitter
- url: https://www.linkedin.com/company/hyperbolic-labs/
type: LinkedIn
- url: https://www.youtube.com/@hyperboliclabs
type: YouTube
- url: https://github.com/HyperbolicLabs
name: Hyperbolic Labs GitHub
type: GitHubOrganization
- url: https://github.com/HyperbolicLabs/Hyper-dOS
name: Hyper-dOS (Distributed Operating System)
type: OpenSourceProject
- url: https://github.com/HyperbolicLabs/Hyperbolic-AgentKit
name: Hyperbolic AgentKit
type: SDK
- url: https://github.com/HyperbolicLabs/hyperbolic-ts
name: Hyperbolic TypeScript Projects
type: SDK
- url: https://github.com/HyperbolicLabs/hyperbolic-mcp
name: Hyperbolic MCP Server
type: Tool
- url: https://github.com/HyperbolicLabs/hyperbolic-cli
name: Hyperbolic CLI
type: Tool
- url: https://github.com/HyperbolicLabs/homebrew-hyperbolic
name: Homebrew Tap for Hyperbolic CLI
type: Tool
- url: https://github.com/HyperbolicLabs/hyperbolic-gradio
name: Hyperbolic Gradio Integration
type: SDK
- url: https://github.com/HyperbolicLabs/hyperbolic-x402
name: Hyperbolic x402 (Coinbase) Chat Completions
type: Tool
- url: https://github.com/HyperbolicLabs/skypilot
name: SkyPilot Fork
type: Tool
- url: https://github.com/HyperbolicLabs/skypilot-catalog
name: SkyPilot Catalog
type: Tool
- url: https://github.com/HyperbolicLabs/jungle.proto
name: Jungle Protocol
type: OpenSourceProject
- url: https://github.com/HyperbolicLabs/inference-benchmarks
name: Inference Benchmarks
type: CodeExamples
- url: https://github.com/HyperbolicLabs/feature-requests
name: Public Feature Requests
type: Forum
- url: https://huggingface.co/docs/inference-providers/en/providers/hyperbolic
name: Hyperbolic on Hugging Face Inference Providers
type: Documentation
- url: https://api.hyperbolic.xyz/v1
name: API Base URL
type: ApiBaseURL
- url: plans/hyperbolic-ai-plans-pricing.yml
type: Plans
- url: rate-limits/hyperbolic-ai-rate-limits.yml
type: RateLimits
- url: finops/hyperbolic-ai-finops.yml
type: FinOps
- url: vocabulary/hyperbolic-ai-vocabulary.yml
type: Vocabulary
- url: rules/hyperbolic-ai-rules.yml
type: SpectralRules
- type: Features
data:
- OpenAI-compatible REST API at https://api.hyperbolic.xyz/v1 — drop-in replacement requiring only api_key
and base_url changes
- 25+ open-source inference models — Llama 3.1 8B/70B/405B, Qwen 2.5, DeepSeek V3, DeepSeek R1, Hermes 3,
Mistral, Llama 3.2 Vision, Qwen2-VL
- Only public provider serving Llama-3.1-405B-Base in BF16 (high-throughput precision) and FP8
(low-latency)
- Image generation with Stable Diffusion XL, SD 3.5, FLUX.1 (sunset), ControlNet, custom LoRAs from
$0.0025/image
- Audio generation (Melo TTS sunset, Whisper coming soon) from $0.001 per 1000 characters
- Decentralized GPU marketplace renting H100, H200, A100, RTX 4090 from third-party suppliers
- On-demand instances deploy in under 1 minute — up to 75% cheaper than AWS / Azure / GCP
- Reserved clusters with volume discounts up to 40% on 3-12 month commitments
- Dedicated single-tenant endpoints for high-throughput inference workloads
- Streaming, tool/function calling, structured JSON output, vision input on supported models
- Python SDK and TypeScript SDK fully OpenAI-compatible
- Hyperbolic CLI (Go) distributed via Homebrew tap
- Model Context Protocol (MCP) server for Claude integration
- Hyperbolic AgentKit — Python agent framework
- Hyper-dOS — open-source Distributed Operating System for GPU orchestration
- Coinbase x402 payment integration for crypto-native chat completions
- Crypto payment support (USDC, USDT, DAI) alongside credit cards
- Hugging Face Inference Providers integration
- Three usage tiers — Basic (60 RPM, free), Pro (600 RPM, $5+ deposit), Enterprise (unlimited, sales)
- 200,000+ builder community
- Marketplace platform fee of 10% on GPU rental income
sources:
- https://www.hyperbolic.ai/
- https://docs.hyperbolic.ai/
- https://docs.hyperbolic.ai/inference/overview
- https://docs.hyperbolic.ai/on-demand/overview
- https://github.com/HyperbolicLabs
- https://huggingface.co/docs/inference-providers/en/providers/hyperbolic
updated: '2026-05-25'
created: '2026-05-25'
modified: '2026-05-25'
position: Consuming
description: Hyperbolic is an open-access AI cloud and decentralized GPU marketplace serving 200,000+ builders
with affordable inference and bare-metal compute. The platform combines a serverless OpenAI-compatible
inference API spanning 25+ open-source LLMs (including the only public Llama-3.1-405B-Base in BF16),
image and audio models, with an on-demand GPU rental marketplace aggregating idle H100 / H200 / A100 /
RTX 4090 capacity from third-party suppliers at 3-10x lower cost than hyperscalers. Reserved clusters,
dedicated endpoints, an OpenAI-drop-in Python and TypeScript SDK, a Go CLI, an MCP server, the Hyperbolic
AgentKit, the open-source Hyper-dOS distributed operating system, and Coinbase x402 crypto payments round
out the stack.
maintainers:
- FN: Kin Lane
email: [email protected]
X: apievangelist
url: https://apievangelist.com
specificationVersion: '0.16'