Zed Pro Costs US$10 and Includes US$5 of Tokens — So What Is the Other US$5 Buying?

🐾 Quick facts

Free tier：There is
Cheapest paid plan：US$10/mo and up
Free quota：Personal: 2,000 accepted edit predictions; the official pricing page …
Last checked：2026-09-21

Article last updated：2026-09-08

One line on Zed’s pricing page looks odd at first: Pro costs US$10 a month and includes US$5 of tokens. Read directly from the official pricing page on 2026-09-08.

So what is the other US$5 for? Here is the breakdown.

1. The three plans, as published

Plan	Price	Edit predictions	Included AI
Personal	US$0 forever	2,000 accepted edit predictions	None — but unlimited use with your API keys or external agents
Pro	US$10/month	Unlimited	US$5 of tokens included
Business	US$30 per seat/month	Unlimited	No fixed allotment — Zed-hosted AI at standard rates, or bring your own keys

The page lists no annual pricing; all three are monthly. Business has no free trial and is cancellable at any time.

2. What the US$10 actually splits into

Cleanly:

US$5 — tokens you can spend
US$5 — the unlimited edit predictions themselves

Edit prediction is Zed’s own completion feature: as you type, it predicts the edit you are about to make. The free plan gives 2,000 accepted predictions — meaning you are only charged when you take one; suggestions you ignore do not count.

So the decision is simple: are you blocked by the 2,000 cap, or by the AI chat allowance?

Blocked on completions → that US$5 is well spent.
Just want more chat → that US$5 buys you none. See the next section.

3. Overage: API list price +10%

Past the included US$5, Zed bills additional token usage at API list price +10%.

Plainly: what costs 100 buying direct costs 110 through Zed. The extra 10% is the hosting margin.

The number is refreshingly honest, and easy to overlook — for heavy users the cost lives in the overage, not in the US$10. Burn US$50 of tokens a month and you land near US$55 through Zed (US$5 included plus roughly US$49.5 for the remaining US$45 at +10%; our arithmetic from Zed’s stated rule, not a published price).

4. The free plan has an unlimited path

The line most easily skipped: Personal supports unlimited use with your own API keys or external agents.

The BYOK list includes Amazon Bedrock, Anthropic, GitHub Copilot, Deepseek, Google AI, LM Studio, Mistral, Ollama, OpenAI, OpenRouter, Vercel and more.

Two genuinely free routes, then:

Bring your own API key — you pay the model provider, with no Zed markup.
Run local models (LM Studio and Ollama are both on the list) — you pay nothing at all.

The one cost is that edit predictions stay capped at 2,000. That is Zed’s own feature; BYOK does not unlock it.

5. One catch in the two-week trial

Pro offers a two-week trial with US$20 of token credits — more than the paid plan’s US$5.

But Zed states that Claude Fable, Claude Opus and GPT Pro are unavailable during the trial.

That matters: if your evaluation is specifically about how the strongest models feel, the trial cannot answer it. What you experience in trial and what you get after paying may be different things.

6. How to choose

Situation	Pick
You want the editor; AI wired up yourself	Personal + own key / local models, US$0
Completions all day, high volume	Pro US$10 — you are buying unlimited predictions
Heavy AI usage, cost-sensitive	Personal + BYOK, avoiding the +10%
Team needing governance and spend visibility	Business US$30/seat

In one line: Zed Pro’s US$10 is mostly buying completions, not AI credits. Once that is clear, the US$5 stops looking strange.

Read directly from zed.dev/pricing on 2026-09-08 and cross-checked against this site’s existing verification record. The US$0/US$10/US$30 prices, 2,000 accepted edit predictions, US$5 included tokens, API list price +10% overage, the two-week US$20 trial excluding Claude Fable/Claude Opus/GPT Pro, and the BYOK provider list are all as published. The US$55 figure in section 3 is our arithmetic from Zed’s stated +10% rule, not a published price. Zed’s page also does not state the reset period for the 2,000 edit predictions.