A 27-endpoint AI compute agent providing seamless LLM inference, image/video generation, speech synthesis, on-chain intelligence, and Bittensor-powered decentralized AI powered by x402 micropayments.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install ai-compute
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install ai-compute using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is AI Compute (ai-compute-x402)?
The ai-compute-x402 skill is a comprehensive, multi-modal compute powerhouse designed to integrate advanced AI capabilities directly into your automation pipelines. As part of the ecosystem of Openclaw Skills, this toolkit provides managed AI operations spanning 11 language models, state-of-the-art image and video generation (FLUX, SDXL), and robust speech-to-text (STT) and text-to-speech (TTS) services. It allows developers to offload complex computational workloads to efficient backend services without heavy local infrastructure.
Beyond traditional centralized compute, the skill offers native routing through Bittensor, a decentralized machine learning network, and integrates specialized on-chain AI intelligence. This enables smart contract auditing, transaction classifications, and wallet profiling directly from your workflows. Every computational transaction is backed by precise x402 micropayments (ranging from $0.001 to $0.50 USDC), complete with a compute-futures prepaid credits system that unlocks volume discounts.
AI Compute (ai-compute-x402) Use Cases
- Multi-Modal Content Generation: Automatically generate marketing materials by chaining text-inference, FLUX-based image generation, and text-to-speech narration into a single fluid pipeline.
- Smart Contract Auditing & On-Chain Intelligence: Conduct automated audits of smart contracts, classify wallet activities, and generate intelligence briefings on decentralized protocols.
- RAG and Knowledge Management: Vectorize extensive document repositories using embeddings, and implement conversational AI agents that query the vector store for responses.
- Decentralized AI Operations: Leverage censorship-resistant, decentralized Bittensor subnets for LLM text inference and image generation.
- Automated Audio Transcriptions: Transcribe voice recordings, audio messages, or meeting logs with speech-to-text, and automatically summarize action items using language models.
How AI Compute (ai-compute-x402) Works
- Authentication & Initialization: The skill securely authenticates with the gateway using the RESEARCH_API_KEY environment variable.
- Endpoint Dispatch: Your agent triggers a command via the
compute.shCLI wrapper, specifying the HTTP method, endpoint, and payload variables. - Micropayment Processing: Each request processes a micropayment through the x402 gateway, debiting from either a direct transaction or a prepaid compute-futures ledger balance.
- Inference Routing: Depending on the endpoint called, the request is directed to centralized clusters (managed GPU servers running Replicate/LLMs) or decentralized nodes via Bittensor subnets.
- Response Delivery: The gateway delivers structured JSON metadata, media links, or classifications back to the client, which can immediately be processed by downstream agents.
AI Compute (ai-compute-x402) Setup
Configure the environment variables required for authorization. Add these to your workspace or deployment configuration to start leveraging Openclaw Skills for compute operations:
# Export the mandatory API key or x402 subscription key
export RESEARCH_API_KEY="your_api_key_here"
# Optional: Customize the gateway URL if using a custom endpoint
export RESEARCH_GATEWAY_URL="https://gateway.spraay.app"
Ensure system dependencies like curl and python3 are globally accessible. You can then trigger any of the 27 endpoints directly from your shell environment:
# Test a text-inference call
bash scripts/compute.sh POST /api/v1/compute/text-inference '{"model":"llama-3.1-8b","prompt":"List three major benefits of decentralized compute.","max_tokens":200}'
AI Compute (ai-compute-x402) Data Schema & Taxonomy
Data Taxonomy and Request Payloads
The ai-compute-x402 skill structures incoming payloads and outgoing telemetry with high-fidelity JSON data types. Below is a taxonomy of the primary schemas utilized across compute operations:
| Category | Parameter | Type | Description |
|---|---|---|---|
| Managed Text | model |
String | Model identifier (e.g., llama-3.1-8b) |
| Managed Text | prompt |
String | Input text context for LLM response generation |
| GPU Inference | input |
Object | Dictionary of custom model parameters forwarded to Replicate |
| On-Chain AI | address |
String | Ethereum/EVM-compatible cryptographic public address |
| Futures | amount |
Number | Amount of USDC to deposit for credit purchasing |
Output Schema Response Structure
All endpoints standardize output structures containing completion metadata, execution cost, and performance telemetry to assist in tracking and accounting:
status: String (success,failed, orpendingfor async tasks)job_id: String (uuid for tracking async jobs)cost: Number (actual micropayment in USDC deducted)data: Object/Array containing inference outputs, token counts, or generated media URLs
AI Compute (ai-compute-x402) Advanced Features
- Volume-Discounted Compute Futures: Deposit USDC credits to lock in tiered discounts, reducing the cost of high-volume inference jobs automatically.
- Batch Processing Pipeline: Dispatch up to 50 compute jobs (mixing text inference, embeddings, and classification) in a single API call to gain a 10% discount on cost.
- Native Decentralized AI (Bittensor): Route queries through the Bittensor network to ensure robust, censorship-resistant inference without single-point failure risks.
- EVM Intelligence Classifiers: Instantly execute transaction classifications, risk audits on smart contracts, and wallet profile generation straight from CLI scripts.
- Custom Replicate Integrations: Run arbitrary public GPU models hosted on Replicate via the streamlined
gpu/runendpoint.
SKILL.md
Loading
Related Openclaw Skills
Energy and Civilization Analyzer
An executable historical and thermodynamic analysis framework that models energy transitions, energy density limits, and EROEI dynamics.
Quantum Computing for Everyone
An executable AI-guided toolkit helping developers and researchers master the core principles of quantum mechanics, qubits, and advanced algorithms.
Arabic UI Localization Reviewer
An AI-powered localization reviewer to audit, refine, and optimize Arabic user interfaces, ensuring flawless RTL layouts and culturally authentic copy.
37°C Soother
A philosophy-psychology soul translator that generates 5 warm, memorable, and comforting quotes to gently soothe the mind on any topic or emotion.
belief-state-runtime
A Python-based epistemic reasoning engine that evaluates claims against retrieved evidence to output structured belief states and calibrated confidence metrics.
Douyin Scraper
An automated data scraper and keyword search utility for Douyin leveraging Playwright and public APIs for seamless social trend retrieval.
Featured*
