AI coding subscriptions, graded: is each plan worth it?
One page per plan: what it costs, what it buys, how far one pass of our fixed task set goes on it, and whether we would pay for it. Measured plans read straight off the provider's meter on the W40 2026 grading run. Plans we did not run on borrow a sibling plan's measurement, scaled by the provider's own advertised multiple and marked est. Two plans side by side is compare subscriptions; a recommendation is the quiz.
| Plan | Provider | Price | Plan usage per pass | Best for |
|---|---|---|---|---|
| Is ChatGPT Free worth it? | OpenAI | Free | not measured | Occasional questions and drafting. |
| Is Cursor Hobby (free) worth it? | Cursor | Free | not measured | Finding out if you want an agent in your editor at all. |
| Is Google AI Plus worth it? | $5/mo | not measured | People who want Gemini inside Gmail and Docs. | |
| Is ChatGPT Go worth it? | OpenAI | $8/mo | not measured | A cheap first step onto Codex. |
| Is SuperGrok Lite worth it? | xAI | $10/mo | not measured | People who already use Grok chat and hit the free limits. |
| Is Kimi Moderato worth it? | Moonshot AI | $19/mo | not measured | Nobody. |
| Is ChatGPT Plus worth it? | OpenAI | $20/mo | 22% of weekly limit GPT-6 Astra | People who want Luna for high-volume small tasks. |
| Is Claude Pro worth it? | Anthropic | $20/mo | ~10% of weekly limitest Claude Opus 5.5 | People who want the best coding subscription at $20. |
| Is Cursor Pro worth it? | Cursor | $20/mo | 2.0% of monthly limit Composer 2.5 | People who still read their code and want to switch between models. |
| Is Google AI Pro worth it? | $20/mo | 23.7% of weekly limit Gemini 3.8 Flash | People who want basic AI usage and will use the Google bundle. | |
| Is SuperGrok worth it? | xAI | $30/mo | not measured | Nobody. |
| Is Kimi Allegretto worth it? | Moonshot AI | $39/mo | 31% of weekly limit Kimi K3 256K | Nobody. |
| Is Kimi Allegro worth it? | Moonshot AI | $99/mo | not measured | Nobody. |
| Is ChatGPT Pro 5x worth it? | OpenAI | $100/mo | ~4.4% of weekly limitest GPT-6 Astra | Codex users blocked by the Plus 5-hour windows. |
| Is Claude Max 5x worth it? | Anthropic | $100/mo | 8% of weekly limit Claude Fable 5.1 | People who keep hitting the Claude Pro limit. Start on Pro first. |
| Is Google AI Ultra 5x worth it? | $100/mo | ~4.7% of weekly limitest Gemini 3.8 Flash | Google AI Pro users blocked by its usage limit. | |
| Is SuperGrok Plus worth it? | xAI | $100/mo | 15% of weekly limit Grok 4.7 | Nobody. |
| Is Kimi Vivace worth it? | Moonshot AI | $199/mo | not measured | Nobody. |
| Is ChatGPT Pro 20x worth it?not available | OpenAI | $200/mo | ~1.1% of weekly limitest GPT-6 Astra | Codex users blocked by the Pro 5x weekly limit. |
| Is Claude Max 20x worth it? | Anthropic | $200/mo | ~4% of weekly limitest Claude Fable 5.1 | People who keep hitting the Max 5x limit. Start on a cheaper plan first. |
| Is Google AI Ultra 20x worth it? | $200/mo | ~1.2% of weekly limitest Gemini 3.8 Flash | People who need maximum Gemini usage and will use the Google bundle. | |
| Is SuperGrok Heavy worth it? | xAI | $300/mo | not measured | Nobody. |
Plan usage is what one pass of our fixed task set (real-world issues plus spec planning, the suites every agent runs alike) consumed of that plan's own allowance, on the best-graded model on the plan. Each figure is a share of a different subscription and never comparable across rows. Estimates scale a sibling plan's measurement by the provider's advertised usage multiple, never by price. Anthropic sells Max as 5x/20x the Pro allowance; our own paired 5x/20x runs measured ~5.7x where 4x is advertised, so the multiple can be well off. Verdicts and prices are the quiz's versioned editorial catalog (v48), last reviewed Oct 1, 2026.