Skip to content
StackLegder Software & AI-tool pricing — verified, dated, and logged when it changes
← All entries
AI assistants & models · guide

AI assistant & model pricing

Updated · 30 comparisons

AI tools are the hardest category to price because the number on the plan is rarely the number that matters. A $20 subscription with a message cap, a $20 subscription with a token budget, and a $20 seat with "fair use" limits are three different products, and the limit is what you actually buy.

The questions the pricing page won't answer directly

What's the real usage limit? Consumer AI subscriptions cap usage — by messages per window, by a monthly credit balance, or by an unpublished rate limit that tightens under load. Two plans at the same price can differ by 5x in how much work you get done before you're throttled.

Is this priced per seat or per usage? Coding assistants and agents increasingly meter requests or "premium" model calls on top of a per-seat fee. A heavy user on a metered plan can cost more than three light users on a flat one.

What happens at the ceiling? Some tools slow you to a weaker model, some queue you, some charge overage, some just stop. That behaviour is part of the price.

Comparing them honestly

Estimate your actual weekly usage — messages, or agent runs, or premium model calls — and place yourself against each tool's limit at the tier you'd buy. The cheapest plan that doesn't throttle you is the real winner, and it's often not the cheapest sticker price. The comparisons below work through the common matchups with the current limits, each verified against the vendor's page and dated.

Comparisons in this guide

Other pricing guides

Every figure on these pages is checked against the provider's own pricing and dated. The complete machine-readable dataset is on the data page.