Now accepting invites — sustainable inference on reclaimed GPUsRequest an invite
COBBLE
Models
Platform
Pricing
Docs
Contact
Sign In
Overview
Getting started
QuickstartMigrate from OpenAIModels
Tool guides
opencodeHermes AgentOpenClawCursor
Reference
API referenceLimits & billingDocuments & embeddings
Documentation

Build on Cobble

Cobble is an OpenAI-compatible inference API. If your code or tools speak to OpenAI, they speak to Cobble — change the base URL and key, and you're running on recycled silicon and renewable energy.

Start the quickstartGet an API key

Getting started

Quickstart

Key → base URL → first request in about two minutes.

Migrate from OpenAI

Drop-in replacement: three changed lines and a model swap.

Models

Discover models programmatically and pick sensible defaults.

Tool guides

opencode

Terminal coding agent — our recommended starting point.

Hermes Agent

Auto-discovers your full Cobble catalog via the wizard.

OpenClaw

Personal AI assistant — provider setup and the allowlist gotcha.

Cursor

Chat-panel integration (read the honest caveat first).

Reference

API reference

Endpoints, streaming, caching, errors, and retry guidance.

Limits & billing

Windows, slots, wallet, and spillover — with real numbers.

Documents & embeddings

OCR, embeddings, and a full PDF → RAG recipe.

Stuck during setup? or contact us.

COBBLE

Built from reclaimed GPUs and server hardware — extending useful life instead of manufacturing new silicon.

Product

  • Models
  • Platform
  • Pricing
  • Docs
  • Status

Company

  • Company
  • Our Story
  • Sustainability
  • News
  • Contact

Legal

  • Terms of Service
  • Privacy Policy
  • Data Processing Addendum
  • Legal Center
© 2026 Cobble Network, Inc. All rights reserved.
Service status