Electricity Bench

Compare subscriptions

Pick two plans and read them side by side: the sourced price, the rung on the provider's own ladder, how the usage window is shaped in our runs, and which graded scorecards stand behind the plan. Prefer a recommendation? Take the quiz; for agent-vs-agent results, compare models.

Choose two plans

OpenAI Free tier

ChatGPT Free

$0/mo per chatgpt.com/pricing, retrieved 2026-08-12

The free ChatGPT tier - limited messages and limited Codex access.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. Same Codex CLI family as ChatGPT Plus with tight limits - the Plus scorecards are the closest evidence.

Cursor Free tier

Cursor Hobby (free)

$0/mo per cursor.com/pricing, retrieved 2026-08-12

A real free tier: the Cursor IDE agent with limited requests.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. We grade the same Cursor agent on the $20 Pro plan - its scorecard is the closest evidence.

Google Entry tier

Google AI Plus

$5/mo per one.google.com, retrieved 2026-08-12

The $5 entry Google plan - Gemini in the chat apps on the cheapest paid rung we track.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. We grade Antigravity CLI on the $20 AI Pro plan - its scorecards are the closest evidence.

OpenAI Entry tier

ChatGPT Go

$8/mo per chatgpt.com/pricing, retrieved 2026-08-12

The $8 entry ChatGPT plan - more messages than free, light Codex access.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. Same Codex CLI family as ChatGPT Plus with tighter limits - the Plus scorecards are the closest evidence.

xAI Entry tier

SuperGrok Lite

$10/mo per x.ai/grok, retrieved 2026-08-12

Basic Grok chat access at $10.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. We grade Grok CLI on the $30 SuperGrok plan - its scorecard is the closest evidence.

Moonshot AI Entry tier

Kimi Moderato

$19/mo per kimi.com/pricing, retrieved 2026-08-12

The $19 Kimi tier with basic agent credits.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. No graded Kimi subject at the moment; see the leaderboard for graded alternatives.

OpenAI Standard tier

ChatGPT Plus

$20/mo per chatgpt.com/pricing, retrieved 2026-08-12

Codex CLI on the $20 mainstream plan; three of our graded agents run on it.

Usage window
Our runs meter it over a weekly window (published run data).
On the bench

Anthropic Standard tier

Claude Pro

$20/mo per claude.com/pricing, retrieved 2026-08-12

The $20 Claude plan; we grade the same models on the bigger Max 20x tier.

Usage window
Same account-window mechanics as Claude Max 20x, where our runs measure a 5-hour rolling window - not a weekly bucket.
On the bench
Not graded on this plan. Same models as Claude Max 20x, smaller allowance - the Max 20x scorecards are the closest evidence.

Cursor Standard tier

Cursor Pro

$20/mo per cursor.com/pricing, retrieved 2026-08-12

The Cursor IDE agent with composer-2.5 at $20 - the plan we grade it on.

Usage window
Our runs meter it over a monthly window (published run data).
On the bench

Google Standard tier

Google AI Pro

$20/mo per one.google.com, retrieved 2026-08-12

Antigravity CLI with Gemini 3.6 Flash and 3.1 Pro at $20 - two graded agents on our leaderboard.

Usage window
Our runs meter it over a weekly window (published run data).
On the bench

xAI Standard tier

SuperGrok

$30/mo per x.ai/grok, retrieved 2026-08-12

Grok CLI with grok-4.5 at $30; the strong mid-price coding option in our runs.

Usage window
Our runs meter it over a weekly window (published run data).
On the bench

Moonshot AI Standard tier

Kimi Allegretto

$39/mo per kimi.com/pricing, retrieved 2026-08-12

Kimi CLI with the open-weight K3 model on the $39 tier - our pick for modest sessions; heavy use belongs on the bigger Kimi tiers.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. The K3 subject sat out the latest weekly grading run, so there is no current scorecard; see the leaderboard for graded alternatives.

Google Heavy tier

Google AI Ultra 5x

$80/mo per one.google.com, retrieved 2026-08-12

More Gemini usage than AI Pro on the $80 Ultra 5x tier.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. Same Antigravity CLI as AI Pro with a bigger allowance - the AI Pro scorecards are the closest evidence.

Moonshot AI Heavy tier

Kimi Allegro

$99/mo per kimi.com/pricing, retrieved 2026-08-12

5x agent credits on the $99 Kimi tier.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. No graded Kimi subject at the moment; see the leaderboard for graded alternatives.

OpenAI Heavy tier

ChatGPT Pro 5x

$100/mo per chatgpt.com/pricing, retrieved 2026-08-12

The $100 Pro tier (5x usage); we grade Codex CLI on the Plus plan.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. Same Codex CLI as ChatGPT Plus with a much bigger allowance - the Plus scorecards are the closest evidence.

Anthropic Heavy tier

Claude Max 5x

$100/mo per claude.com/pricing, retrieved 2026-08-12

The $100 middle Max tier - the same Claude family we grade on Max 20x, with a smaller window.

Usage window
Same account-window mechanics as Claude Max 20x, where our runs measure a 5-hour rolling window - not a weekly bucket.
On the bench
Not graded on this plan. Same models and harness as Claude Max 20x on a quarter of the allowance - the Max 20x scorecards are the closest evidence.

xAI Heavy tier

SuperGrok Plus

$100/mo per x.ai/grok, retrieved 2026-08-12

Significantly higher Grok usage on the $100 tier.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. Same Grok CLI as SuperGrok with a bigger allowance - the SuperGrok scorecard is the closest evidence.

Google Top tier

Google AI Ultra 20x

$180/mo per one.google.com, retrieved 2026-08-12

The $180 top Google tier (Ultra 20x).

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. Same Antigravity CLI as AI Pro with a far bigger allowance - the AI Pro scorecards are the closest evidence.

Moonshot AI Top tier

Kimi Vivace

$199/mo per kimi.com/pricing, retrieved 2026-08-12

10x agent credits on the $199 top Kimi tier.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. No graded Kimi subject at the moment; see the leaderboard for graded alternatives.

OpenAI Top tier

ChatGPT Pro 20x

$200/mo per chatgpt.com/pricing, retrieved 2026-08-12

The top Pro tier (20x usage); we grade Codex CLI on the Plus plan.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. Same Codex CLI as ChatGPT Plus with a far bigger allowance - the Plus scorecards are the closest evidence.

xAI Top tier

SuperGrok Heavy

$300/mo per x.ai/grok, retrieved 2026-08-12

The $300 top xAI tier with the highest usage and fastest speed.

Usage window
No published quota readout for this plan.
On the bench
Not graded on this plan. Same Grok CLI as SuperGrok with a far bigger allowance - the SuperGrok scorecard is the closest evidence.

Prices and plan facts are the quiz's versioned editorial catalog (v18), reviewed in git - never computed from the live leaderboard. Quota shares stay a share of each plan's own allowance, so this page never subtracts one plan's usage figure from another's; the honest cross-plan cost comparison lives on the leaderboard as work per plan-dollar.