Which plan can handle my work?
How is that spread over the day?
What do you mostly do?
Which models do you use?Default: $0.08 per request
We assume a mid-range coding model at $0.08 per request. Save your own models in My setup and the answer will use their prices.
Models change only the API cost of your work, not whether a plan's limits hold.
Open My setupHow the check works
Count the hours when a model is actually busy, not the whole working day.
Long sessions, then quiet — a 5-hour window fills 1.5 times faster.
An agent edits code and runs tools for you — about 12 long requests an hour.
How the check worksCheapest plan that covers this work
Covers your work with room · by the vendor's own promise only — less certain
- Share used
- 31%
- Same work through the API
- $173 a month
Calculation and range
Uses about 31% of the plan (between 6% and 56%).
Through the API the same work would cost $173 — 6.2 times this plan's price.
Alternative backed by limits or measurement
by the vendor's published limits
Covers it with room11
- Kimi Allegretto$28 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 31% of the plan (between 6% and 56%).
Through the API the same work would cost $173 — 6.2 times this plan's price.
- Qwen Token Pro$70 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 15% of the plan (between 3% and 28%).
Through the API the same work would cost $173 — 2.5 times this plan's price.
- GLM Coding Pro$72 a month
Published limits
by the vendor's published limits
Uses about 25% of the plan (between 23% and 28%). Of its limits, the weekly cap is the closest to its edge.
Through the API the same work would cost $173 — 2.4 times this plan's price.
- Kimi Allegro$98 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 3% of the plan (between 1% and 5%).
Through the API the same work would cost $173 — 1.8 times this plan's price.
- Claude Max 5x$100 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 35% of the plan (between 7% and 62%).
Through the API the same work would cost $173 — 1.7 times this plan's price.
- GitHub Copilot Max$100 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 60% of the plan (between 24% and 95%).
Through the API the same work would cost $173 — 1.7 times this plan's price.
- Trae Ultra$100 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 9% of the plan (between 3% and 14%).
Through the API the same work would cost $173 — 1.7 times this plan's price.
- GLM Coding Max$160 a month
Published limits
by the vendor's published limits
Uses about 6% of the plan (between 6% and 7%). Of its limits, the weekly cap is the closest to its edge.
Through the API the same work would cost $173 — 1.1 times this plan's price.
- Claude Max 20x$200 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 4% of the plan (between 1% and 8%).
The API would be cheaper: $173 against $200 for the plan.
- Cursor Ultra$200 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 4% of the plan (between 1% and 8%).
The API would be cheaper: $173 against $200 for the plan.
- Google AI Ultra$200 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 4% of the plan (between 1% and 8%).
The API would be cheaper: $173 against $200 for the plan.
Runs out7
- Google AI Plus$4.99 a month
Vendor promise
by the vendor's own promise only — less certain
Needs about 17 times what the plan allows — at least 3.5 times.
Through the API the same work would cost $173 — 35 times this plan's price.
- Kimi Moderato$14 a month
Vendor promise
by the vendor's own promise only — less certain
Needs about 3.1 times what the plan allows.
Through the API the same work would cost $173 — 12 times this plan's price.
- GLM Coding Lite$18 a month
Published limits
by the vendor's published limits
Uses about 126% of the plan (between 113% and 139%). The weekly cap runs out first.
Through the API the same work would cost $173 — 9.6 times this plan's price.
- Qwen Token Standard$19.60 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 220% of the plan (from 44% to over 300%).
Through the API the same work would cost $173 — 8.8 times this plan's price.
- Google AI Pro$19.99 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 216% of the plan (from 43% to over 300%).
Through the API the same work would cost $173 — 8.6 times this plan's price.
- Trae Pro+$30 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 165% of the plan (between 66% and 263%).
Through the API the same work would cost $173 — 5.8 times this plan's price.
- GitHub Copilot Pro+$39 a month
Vendor promise
by the vendor's own promise only — less certain
Uses about 111% of the plan (between 44% and 177%).
Through the API the same work would cost $173 — 4.4 times this plan's price.
Can't tell34
These plans publish neither limits nor a usage multiplier, so nothing honest can be said about them.
Show the 34 plans
Qwen Token Lite, MiMo Lite, Kimi Andante, Meta One Plus, Doubao Standard, GitHub Copilot Pro, Trae Pro, Le Chat Pro, MiMo Standard, Meta One Premium, Microsoft 365 Premium, ChatGPT Plus, Claude Pro, Cursor Pro, Devin Pro, MiniMax Plus, Perplexity Pro, Lovable Pro, Replit Core, SuperGrok, v0 Plus, Lovable Business, MiMo Pro, MiniMax Max, Cursor Pro+, Doubao Professional, MiMo Max, Replit Pro, v0 Business, MiniMax Ultra, ChatGPT Pro, Devin Max, Perplexity Max, SuperGrok Heavy
What this work costs through the API$173a month
About 2200 requests a month at $0.08 per request. Busiest 5 hours: 90 requests; a week: 504.
Cheaper than the API: a flat plan inside a coding agent
Z.ai officially lets you run GLM-5.2 (Max) inside Claude Code. Through the API your work costs about $173 a month; the GLM Coding Pro plan is $72 flat — 42% of that.
Z.ai claims the plan covers $1080–2160 a month of API value. Nobody has measured it yet. Your work fits even the cautious end of that.
Covers your work with room (by the vendor's published limits)