ClawRouter | Awesome Repository

ClawRouter is an AI model router and API gateway designed to classify query complexity and assign prompts to the most efficient model tier. It operates as a multi-model AI proxy that orchestrates traffic between various large language models and AI media generators through a unified interface.

The project distinguishes itself by integrating a non-custodial micropayment processor using the x402 protocol. This allows for per-request API access and USDC settlement on Base and Solana chains, replacing static API keys with wallet-based authentication and real-time budget enforcement.

The system covers broad capability areas including automated model failover management, response caching for cost control, and multi-modal proxying for text, image, and video generation. It further integrates blockchain data services, enabling on-chain SQL querying, financial market data retrieval, and wallet intelligence analysis.

The project is implemented as a local AI proxy server that intercepts and routes standardized AI API requests to various backends.

Features

Model Routers - Acts as a model router that dynamically assigns prompts to the most efficient model tier based on query complexity.

Micropayment Protocols - Implements the x402 protocol for non-custodial USDC micropayments to settle per-request AI API fees.

Agentic Task Detection - Identifies multi-step or tool-using requests via a scoring pathway to trigger specialized model configurations.

Model Tier Selectors - Implements a logic for selecting model tiers based on query complexity to optimize the balance between cost and quality.

LLM Fallback Managers - Controls traffic by rerouting failed requests across a chain of alternative providers to ensure continuous availability.

LLM Gateways - Provides a centralized gateway that aggregates multiple LLM providers into a single standardized API.

LLM Response Streaming - Converts non-streaming JSON API responses into Server-Sent Events for token-by-token delivery.

Capability-Based Routing - Implements routing logic that selects AI models based on specific functional capabilities like vision processing.

Failover Mechanisms - Implements automatic fallback chains to switch AI model providers when primary services experience outages.

Cost-Performance Optimizers - Optimizes costs and performance by routing requests to the most efficient model tier based on complexity.

Model Fallbacks - Provides a prioritized chain of alternative models to maintain availability when primary providers return errors.

Multi-Model AI Orchestrators - Operates as an orchestrator that connects diverse AI models and media generators via a unified local interface.

Prompt Complexity Classification - Classifies prompt complexity by analyzing reasoning, code, and technical requirements to assign a difficulty tier.

Blockchain-Based Settlements - Settles API usage fees using USDC on Solana and Base chains with integrated wallet verification.

Non-Custodial Payment Flows - Handles service authentication and payments via non-custodial USDC transactions on Base or Solana.

Micropayment Protocol Implementations - Implements the x402 protocol for executing per-request micropayments on blockchain networks.

Response Caching - Caches model responses locally to serve repeat requests instantly and reduce API costs.

AI Provider Proxies - Implements a local proxy server designed to route and manage traffic to multiple large language model providers.

Server-Sent Events - Converts standard JSON API responses into an incremental event stream for real-time content delivery.

Crypto Wallet Content Payments - Signs payment payloads locally using wallet keys to comply with micropayment protocols for model access.

x402 Micropayments - Uses the x402 protocol and non-custodial wallets to handle per-request USDC micropayments for AI API usage.

On-Chain Payment Verification - Monitors USDC balances across EVM and Solana networks to verify liquidity before executing model calls.

Wallet Signature Authentication - Uses cryptographic wallet signatures to authorize per-request API access and replace static API keys.

Payment-Gated Access - Restricts API access and processes micropayments using wallet authentication and the x402 standard.

Task Complexity Analyzers - Analyzes prompt reasoning and technical requirements across weighted dimensions to assign complexity tiers.

Complexity-Based Routers - Automatically assigns the most cost-effective model tier by analyzing query complexity through weighted scoring.

API Request Deduplication - Prevents duplicate billing by hashing request bodies to identify and replay cached responses.

Complexity Tier Mappings - Maps prompt complexity tiers to specific models based on cost, performance, and quality optimization goals.

AI Image Generation - Integrates AI models to produce visual assets from text prompts with configurable output sizes.

AI Video Generators - Synthesizes short videos from text or image inputs and proxies the resulting media to local storage.

Dynamic Tier Escalation - Upgrades the model complexity tier automatically after detecting repeated identical requests to break agentic loops.

Image Editing - Enables modification of existing images using text prompts and masks via generative AI instructions.

Model Performance Evaluators - Evaluates language models using latency and pricing metrics to optimize routing efficiency.

Provider Failover Handlers - Maps provider-specific errors to standardized categories to manage retries and provider failover.

Web Search Integrations - Connects the system to external search providers to enable augmented real-time web capabilities for models.

AI Token Budget Controllers - Enforces strict spending limits and per-session cost ceilings to prevent billing overages on AI model requests.

AI Agent Payment Frameworks - Provides a unified system for managing model usage costs and micropayment settlement across multiple providers.

Prediction Markets - Fetches market rankings and trade volumes from prediction platforms via a unified interface.

Real-Time Usage Tracking - Tracks real-time consumption of billable units via response headers to enforce budget caps.

Blockchain Data Analytics - Extracts and queries historical ledger records, market data, and token analytics via a unified routing layer.

Crypto Data Routing - Forwards requests to a unified crypto data API covering exchange prices, market rankings, and project metrics.

LLM Token Compression - Reduces billed token counts by deduplicating messages and minifying data before sending requests.

Request Body Hashing - Identifies duplicate requests by hashing payloads to serve cached responses and prevent redundant billing.

On-Chain SQL Execution - Executes structured SQL queries against on-chain data and retrieves schema definitions for introspection.

Wallet Intelligence Analysis - Retrieves detailed wallet profiles, net worth, and transaction history for specific blockchain addresses.

Rate Limiters - Tracks rate limits on a per-model basis to prevent provider failures from blocking fallback chains.

Multi-Modal Provider Proxies - Routes requests for text, image, and video generation through a single interface to unify different provider APIs.

Cost Estimation & Buffering - Estimates request costs and signs payment buffers in advance to bypass the payment-required round trip.

Routing Diagnostics - Provides technical diagnostics on model selection, session state, and routing scores to troubleshoot failures.

AI Reasoning Inspections - Attaches debug headers to responses to expose the chosen model, confidence scores, and routing reasoning.

Model Tool Calls - Converts textual or XML tool invocations into structured tool calls for downstream API execution.