How do I reduce my Cursor API bill?

Cursor Pro's $20/mo covers 500 fast requests — past that you pay OpenAI / Anthropic per-call rates directly, which is why heavy users end up at $50-$200/mo. The fix is to point Cursor's Custom API at CodeRouter and set model to 'auto'. CodeRouter detects what phase of coding the request is (planning, implementation, debugging, test generation, docs) and routes to the cheapest capable model per phase — Opus only for planning, DeepSeek V3 for test generation, Haiku for docstrings. Same Cursor IDE, same keyboard shortcuts, 70-90% lower monthly bill. Setup takes 2 minutes — just change base_url to https://www.coderouter.io/api/v1 and paste your cr_ API key.

What is the cheapest API for Claude Code / Aider / Copilot?

There isn't a single 'cheapest API' — the cheapest model depends on what the coding agent is doing. For planning and architecture, you still want Claude Opus 4.7 or Sonnet 4.6. For implementation, DeepSeek V3.2 ($0.28/$0.42 per 1M) and Qwen 3 Coder are 30-50x cheaper than Opus with near-equivalent code quality. For test generation, DeepSeek V3 or Haiku 4.5 is 15-50x cheaper. For docstrings and simple formatting, Haiku 4.5 ($1/$5) or Gemini 2.5 Flash ($0.30/M output) is 15-250x cheaper. CodeRouter is the gateway that picks per request automatically — aim a single base_url at https://www.coderouter.io/api/v1 from Claude Code, Aider, Copilot (via LiteLLM), Cursor, Windsurf, or any OpenAI-compatible agent.

Does DeepSeek V3 work as well as Claude Sonnet for coding?

For implementation and test generation phases — yes, DeepSeek V3.2 matches Claude Sonnet 4.6 on HumanEval, MBPP, and LiveCodeBench within 1-3 points. For multi-file refactoring and architecture planning, Sonnet still has an edge on long-context reasoning. DeepSeek V3 costs $0.28 input / $0.42 output per 1M tokens, vs Sonnet's $3/$15 — roughly 30-50x cheaper. The right answer for most coding agents is not 'pick one forever' but 'use DeepSeek V3 for the implement/test phases and Sonnet for the plan/refactor phases.' That's phase-aware routing in practice — CodeRouter decides per request in ~10ms.

What is phase-aware LLM routing?

Phase-aware LLM routing classifies each coding-agent request by what phase of software work it represents — planning, implementation, debugging, testing, refactoring, or documentation — and routes it to the cheapest model that can handle that specific phase. A 'write unit tests for this function' request goes to DeepSeek V3 ($0.42/M). A 'refactor this multi-file feature and plan the migration' request goes to Claude Opus 4.7 ($75/M). This is different from picking one model for everything, and different from OpenRouter-style model-selection (which still requires you to choose manually). CodeRouter's classifier runs in ~10ms on the server, so the agent never notices the extra hop.

CodeRouter vs OpenRouter — which saves more money on coding?

OpenRouter is a model marketplace — it gives you access to 300+ models behind one API key, but you still pick which model to send each request to. Most Cursor / Aider / Claude Code users default to the premium model (Opus, GPT-5) for everything and end up paying full price. CodeRouter is a phase-aware router — set model to 'auto' and we pick the cheapest capable model per request based on the coding phase. CodeRouter also adds things OpenRouter doesn't: coding-specific capability scores per model (implementation, debug, test, refactor), per-end-user attribution for SaaS agent builders, and built-in quota + top-up billing. For pure coding workloads, typical CodeRouter savings are 70-90% vs picking one model on OpenRouter.

Will CodeRouter break my Cursor / Aider / Claude Code agent?

No. CodeRouter exposes a standard OpenAI-compatible chat completions endpoint (POST /api/v1/chat/completions) with the same request and response format your agent already uses — including streaming, tool use, and function calling. We implement the same JSON schema and stream format, so Cursor, Aider, Claude Code, Cline, Continue.dev, Windsurf, OpenClaw, and any LiteLLM-wrapped client work unmodified. If a routed model fails, the fallback chain tries up to 2 alternates automatically (on 429, 500-504, timeouts, missing keys). You can also pin an explicit model instead of 'auto' any time.

How do I set up CodeRouter with Cursor in 2 minutes?

1) Sign up free at https://www.coderouter.io/login and copy your API key (starts with cr_). 2) In Cursor, open Settings -> Models -> OpenAI API Key, and under 'Override OpenAI Base URL' paste https://www.coderouter.io/api/v1. Paste your cr_ key in the API Key field. 3) Add 'auto' to the Custom Models list and select it as your active model. That's it — phase-aware routing is live. Aider users set OPENAI_API_BASE and OPENAI_API_KEY env vars to the same values. Claude Code users set ANTHROPIC_BASE_URL to https://www.coderouter.io/api/v1 and ANTHROPIC_API_KEY to the cr_ key. Full guide at https://www.coderouter.io/setup.

Setup Guide

Get CodeRouter running with your AI agent in under 60 seconds.

Pick a plan

All paid plans use system provider keys; the Free Trial gives you 1M tokens (over 2 days) to evaluate routing quality.

Starter / Solo — side projects and indie devs ($19–$29/mo).
Pro / Studio / Team — heavy agent workflows or small teams ($99–$299/mo, up to 50 keys).
Free Trial — no card needed, 1M tokens / 2 days to evaluate.

See plans

Get your API key

Sign in with Google, then create a key in the dashboard. Name it after the engineer or workflow (e.g. alice, ci, prod) so per-key analytics make sense later.

Configure your AI agent

One-line setup script that detects your installed agent (OpenClaw, Cursor, Aider, Claude Code) and writes the right config:

Anthropic's CLI. Uses ANTHROPIC_AUTH_TOKEN — NOT ANTHROPIC_API_KEY (that one only works against api.anthropic.com directly).

One-shot env vars

export ANTHROPIC_BASE_URL=https://www.coderouter.io/api/v1
export ANTHROPIC_AUTH_TOKEN=cr_YOUR_KEY_HERE
export ANTHROPIC_MODEL=auto
export ANTHROPIC_DEFAULT_OPUS_MODEL=auto
export ANTHROPIC_DEFAULT_SONNET_MODEL=auto
export ANTHROPIC_DEFAULT_HAIKU_MODEL=auto
# Optional v2.x fix: disable experimental beta features that
# some third-party gateways don't yet handle.
export CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1
claude

~/.claude/settings.json (persistent)

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://www.coderouter.io/api/v1",
    "ANTHROPIC_AUTH_TOKEN": "cr_YOUR_KEY_HERE",
    "ANTHROPIC_MODEL": "auto",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "auto",
    "ANTHROPIC_DEFAULT_SONNET_MODEL": "auto",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "auto",
    "CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS": "1"
  }
}

⚠ Gotcha: If the badge in Claude Code's header shows 'Claude Max' instead of 'API Usage Billing' after restart, your existing OAuth session is winning over the env vars. Run `claude logout`, then relaunch.

Need a different agent (Cline, Windsurf, GitHub Copilot via LiteLLM)? See full integration docs.

Verify it works

Test routing without burning tokens — X-Dry-Run returns the routing decision (model, phase, cost estimate) without calling the provider:

Dry-run test

curl https://www.coderouter.io/api/v1/chat/completions \
  -H "Authorization: Bearer cr_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -H "X-Dry-Run: true" \
  -d '{"model":"auto","messages":[{"role":"user","content":"Hello"}]}'

Response should include the chosen model and phase. Drop X-Dry-Run for a real call.

Need Help?

Join our community or reach out directly.

Contact Support