How many tokens does an 8-point story take to build with AI?
A 8-point user story typically takes about 1,680,000 tokens to build with an AI coding agent (roughly 735,000–3,680,000 depending on complexity), about $4.80 at today's GPT-4o prices ($2.10–$10.55). This is the build-time (developer) cost, not the cost of running the finished feature.
An 8-point story is a large chunk of work — cross-cutting changes across several components with lots of iteration. Many teams treat 8 as 'consider splitting this,' and for AI that shows up as millions of input tokens over the session.
Token & cost ranges
Cost uses GPT-4o at current prices. Verified 29 July 2026.
| Scenario | Input tokens | Output tokens | Total tokens | Est. cost |
|---|---|---|---|---|
| Low (simple) | 700,000 | 35,000 | 735,000 | $2.10 |
| Typical | 1,600,000 | 80,000 | 1,680,000 | $4.80 |
| High (complex) | 3,500,000 | 180,000 | 3,680,000 | $10.55 |
Typical cost by model
The same typical token estimate priced across a few common coding models.
| Model | Typical cost | Range (low–high) |
|---|---|---|
| GPT-4o | $4.80 | $2.10 – $10.55 |
| GPT-4o mini | $0.2880 | $0.1260 – $0.6330 |
| Claude Sonnet 4.6 | $6.00 | $2.62 – $13.20 |
| Gemini 2.0 Flash | $0.1920 | $0.0840 – $0.4220 |
How we estimate this
An 8-point story is large — cross-cutting changes, several components, and many iterations, often spanning more than a day. Input tokens dominate because an agent re-sends file context on every turn; output is the generated code, diffs and tests. Treat these as planning ranges, not a quote — your real usage depends on codebase size, model, prompt caching and how many iterations it takes.
Examples of this size
- A feature touching frontend, backend and schema
- A non-trivial integration with error handling
- A subsystem rewrite with a full test suite
Estimate it with your own model
Plug in your model, complexity and prompt-caching to get a tailored range.
Frequently asked questions
How many tokens does a 8-point user story take to build with AI?+
A 8-point user story typically takes about 1,680,000 tokens to build with an AI coding agent (roughly 735,000–3,680,000 depending on complexity), about $4.80 at today's GPT-4o prices ($2.10–$10.55). This is the build-time (developer) cost, not the cost of running the finished feature.
Is this the cost to build it or to run it?+
This is the build-time (developer) cost — the tokens an AI coding agent burns while implementing the work. Running the finished feature is a separate, usually much smaller per-request cost.
Why is the range so wide?+
An 8-point story is large — cross-cutting changes, several components, and many iterations, often spanning more than a day. Input tokens dominate because an agent re-sends file context on every turn; output is the generated code, diffs and tests. Treat these as planning ranges, not a quote — your real usage depends on codebase size, model, prompt caching and how many iterations it takes.
How can I lower the cost?+
Split large work into smaller pieces, enable prompt caching so re-sent context is billed at the cheaper cache-read rate, and use a cheaper model for routine steps. Try your own model in the Cost to build X estimator.