Up to 95% Off Cached Input
Cursor frequently re-sends your codebase context. Newton’s passes Anthropic’s prompt-cache pricing straight through — cached input tokens cost up to 95% less than fresh ones, and the saving shows on every receipt.
IDE INTEGRATION GUIDE
Hit your monthly Cursor rate limits or tired of paying for idle seats? Route Cursor through Newton’s to unlock pay-as-you-go Claude models, automatic prompt-caching discounts, and edge-fast first token responses.
Save money and skip rate limits on intensive AI coding workflows.
Cursor frequently re-sends your codebase context. Newton’s passes Anthropic’s prompt-cache pricing straight through — cached input tokens cost up to 95% less than fresh ones, and the saving shows on every receipt.
Never get locked out mid-sprint. Pay only for exact token usage from a unified prepaid balance with zero recurring monthly lock-in.
Switch between Claude Opus 5.5 for the hardest problems, Sonnet 5.5 for everyday coding, and Haiku 4.5 for fast edits — all on one key, with no separate API accounts to manage.
No extensions needed. Use Cursor’s built-in custom OpenAI provider settings.
Press Cmd + , (Mac) or Ctrl + , (Windows/Linux) and navigate to Cursor Settings > Models.
Turn on OpenAI API Key, toggle Override OpenAI Base URL, and enter:
Under Model Names, add live verified routes from the Newton’s catalog:
claude-sonnet-5-5 (Live & Verified — the default: frontier everyday coding & agentic workflows)claude-opus-5-5 (Live & Verified — hardest problems & deep reasoning)claude-haiku-4-5-20251001 (Live & Verified — fastest responses for quick edits & debugging)claude-opus-5 / claude-sonnet-5 / claude-sonnet-4-6 / claude-opus-4-8 / claude-opus-4-7 / claude-opus-4-6 (Also live — see the full catalog in /models)Open Cursor Composer (Cmd + I / Ctrl + I) and send a quick prompt. Watch your request log instantly in your Newton’s console with an itemized token receipt!
Bypass Cursor Pro monthly fast request exhaustion and upstream provider throttling with Newton's prepaid balance, published per-token pricing, and an itemized receipt for every request.
Read Cursor Rate Limit Troubleshooting Guide →Make your agent faster, more accurate, and consume 30% fewer tokens with standardized project rules.