Inference Registry · Est. 2026

InferenceforAgents

Subsidized tokens for agents. Pay with or .

How It Works

Frontier inference at half price

AI subscription plans like Claude Code Max and ChatGPT Pro provide 10-30x more compute than what you pay for at raw API rates. Autotheft leverages this subsidy to offer frontier model inference at 50% off official API pricing.

Subsidized Compute

Subscription plans bundle massive compute at flat rates. We pass those savings directly to agents at half the official per-token price.

Crypto-Native Payments

Pay with or . Deposit once for bulk credits, or pay per-request with x402. No credit cards, no KYC.

OpenAI-Compatible

Drop-in replacement for any OpenAI SDK integration. Just change the base URL and API key — same request format, same response schema.

$base_url=autotheft.dev/api/v1

Model Registry

Models

Frontier AI at a fraction of the cost.

OpenAINew

GPT-5.5

Flagship reasoning model for complex coding

$2.50/ 1M input$5.00

AnthropicPopular

Claude Opus 4.8

Most capable reasoning & agentic coding model

$2.50/ 1M input$5.00

AnthropicFast

Claude Haiku 4.5

Fastest, near-frontier intelligence

$0.50/ 1M input$1.00

GoogleCheapest

Gemini 3.1 Flash-Lite

Ultra-fast, lowest cost

$0.13/ 1M input$0.25

Anthropic

Claude Sonnet 4.6

Best balance of speed and intelligence

$1.50/ 1M input$3.00

OpenAI

GPT-5.4 mini

Fast, low-cost coding and computer use

$0.38/ 1M input$0.75

OpenAI

GPT-5.4

Balanced model for production coding

$1.25/ 1M input$2.50

Google

Gemini 3.5 Flash

Frontier intelligence with Flash-class speed

$0.75/ 1M input$1.50

Google

Gemini 3.1 Pro

Google's most capable reasoning model

$1.00/ 1M input$2.00

OpenAINew

GPT-5.5

Flagship reasoning model for complex coding

$2.50/ 1M input$5.00

AnthropicPopular

Claude Opus 4.8

Most capable reasoning & agentic coding model

$2.50/ 1M input$5.00

AnthropicFast

Claude Haiku 4.5

Fastest, near-frontier intelligence

$0.50/ 1M input$1.00

GoogleCheapest

Gemini 3.1 Flash-Lite

Ultra-fast, lowest cost

$0.13/ 1M input$0.25

Anthropic

Claude Sonnet 4.6

Best balance of speed and intelligence

$1.50/ 1M input$3.00

OpenAI

GPT-5.4 mini

Fast, low-cost coding and computer use

$0.38/ 1M input$0.75

OpenAI

GPT-5.4

Balanced model for production coding

$1.25/ 1M input$2.50

Google

Gemini 3.5 Flash

Frontier intelligence with Flash-class speed

$0.75/ 1M input$1.50

Google

Gemini 3.1 Pro

Google's most capable reasoning model

$1.00/ 1M input$2.00

Google

Gemini 3.5 Flash

Frontier intelligence with Flash-class speed

$0.75/ 1M input$1.50

Google

Gemini 3.1 Pro

Google's most capable reasoning model

$1.00/ 1M input$2.00

AnthropicFast

Claude Haiku 4.5

Fastest, near-frontier intelligence

$0.50/ 1M input$1.00

OpenAI

GPT-5.4

Balanced model for production coding

$1.25/ 1M input$2.50

Anthropic

Claude Sonnet 4.6

Best balance of speed and intelligence

$1.50/ 1M input$3.00

OpenAINew

GPT-5.5

Flagship reasoning model for complex coding

$2.50/ 1M input$5.00

OpenAI

GPT-5.4 mini

Fast, low-cost coding and computer use

$0.38/ 1M input$0.75

GoogleCheapest

Gemini 3.1 Flash-Lite

Ultra-fast, lowest cost

$0.13/ 1M input$0.25

AnthropicPopular

Claude Opus 4.8

Most capable reasoning & agentic coding model

$2.50/ 1M input$5.00

Google

Gemini 3.5 Flash

Frontier intelligence with Flash-class speed

$0.75/ 1M input$1.50

Google

Gemini 3.1 Pro

Google's most capable reasoning model

$1.00/ 1M input$2.00

AnthropicFast

Claude Haiku 4.5

Fastest, near-frontier intelligence

$0.50/ 1M input$1.00

OpenAI

GPT-5.4

Balanced model for production coding

$1.25/ 1M input$2.50

Anthropic

Claude Sonnet 4.6

Best balance of speed and intelligence

$1.50/ 1M input$3.00

OpenAINew

GPT-5.5

Flagship reasoning model for complex coding

$2.50/ 1M input$5.00

OpenAI

GPT-5.4 mini

Fast, low-cost coding and computer use

$0.38/ 1M input$0.75

GoogleCheapest

Gemini 3.1 Flash-Lite

Ultra-fast, lowest cost

$0.13/ 1M input$0.25

AnthropicPopular

Claude Opus 4.8

Most capable reasoning & agentic coding model

$2.50/ 1M input$5.00

InfrastructureNo. 01

Agent-Native Infrastructure

Built from the ground up for autonomous agents. No human-oriented dashboards or manual approvals in the critical path.

Programmatic registration and key management
OpenAI-compatible API — drop-in replacement
50% subsidized pricing on all frontier models
Sub-second latency via optimized routing

# Agent registers itself

const res = await fetch(

"/api/v1/register",

{ method: "POST" }

);

const { api_key, deposit_address } = await res.json();

PaymentsNo. 02

x402 — Zero Friction Payments

The x402 protocol enables machine-to-machine payments over HTTP. No pre-registration, no API keys — just a wallet.

HTTP-native payment protocol
No registration or API key required
Automatic 402 → pay → 200 flow
Works with any x402-compatible wallet

x402 flow

POST /api/v1/chat/completions

402 Payment Required → signs tx

200 OK → completion stream

The Process

How It Works

Three steps. No credit cards, no KYC, no human in the loop.

No.01

Register

Get your API key in a single POST call. No dashboards, no waiting.

terminal

POST /api/v1/register
// => { api_key: "at_..." }

No.02

Deposit

Fund your account on Solana or Base.

terminal

SEND ◉ | ◉
// => { balance: "$50.00" }

No.03

Call

Drop-in OpenAI-compatible API. Same format, half the price.

terminal

POST /v1/chat/completions
// => { choices: [...] }

Price Sheet

Pricing

Pay less. Get more.

ModelOur InputOfficial InputOur OutputOfficial OutputSavings

Claude Opus 4.8Popular$2.50$5.00$12.50$25.000%

Claude Sonnet 4.6$1.50$3.00$7.50$15.000%

Claude Haiku 4.5Fast$0.50$1.00$2.50$5.000%

GPT-5.5New$2.50$5.00$15.00$30.000%

GPT-5.4$1.25$2.50$7.50$15.000%

GPT-5.4 mini$0.38$0.75$2.25$4.500%

Gemini 3.1 Pro$1.00$2.00$6.00$12.000%

Gemini 3.5 Flash$0.75$1.50$4.50$9.000%

Gemini 3.1 Flash-LiteCheapest$0.13$0.25$0.75$1.500%

Claude Opus 4.8Popular0%

Input

$2.50 $5.00

Output

$12.50 $25.00

Claude Sonnet 4.60%

Input

$1.50 $3.00

Output

$7.50 $15.00

Claude Haiku 4.5Fast0%

Input

$0.50 $1.00

Output

$2.50 $5.00

GPT-5.5New0%

Input

$2.50 $5.00

Output

$15.00 $30.00

GPT-5.40%

Input

$1.25 $2.50

Output

$7.50 $15.00

GPT-5.4 mini0%

Input

$0.38 $0.75

Output

$2.25 $4.50

Gemini 3.1 Pro0%

Input

$1.00 $2.00

Output

$6.00 $12.00

Gemini 3.5 Flash0%

Input

$0.75 $1.50

Output

$4.50 $9.00

Gemini 3.1 Flash-LiteCheapest0%

Input

$0.13 $0.25

Output

$0.75 $1.50

No. 402 — Settlement

Payment Rails

Two ways to pay. Zero friction.

Zero SetupNo. 01

x402 — Pay Per Request

No registration. No API keys. Just in your wallet.

SolanaBase

const x402Fetch = wrapFetch(fetch, signer);

const res = await x402Fetch(

"https://autotheft.dev/api/v1/chat/completions"

);

Bulk DiscountNo. 02

Prepaid Credits

SOLUSDC

curl -X POST /api/v1/register \

-d '{"name": "my-agent"}'

# → { "api_key": "at_...", "deposit_address": "So..." }

No. 010 models

Models Available

No. 020 chains

Chains Supported

No. 030 rails

Payment Rails