How many tokens does a 2-point story take to build with AI?
A 2-point user story typically takes about 318,000 tokens to build with an AI coding agent (roughly 128,000–690,000 depending on complexity), about $0.9300 at today's GPT-4o prices ($0.3800–$2.03). This is the build-time (developer) cost, not the cost of running the finished feature.
A 2-point story is still small, but there's a bit more to wire up than a 1-pointer. Expect a few agent turns across a couple of files, so token usage roughly doubles versus a 1-point change.
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) | 120,000 | 8,000 | 128,000 | $0.3800 |
| Typical | 300,000 | 18,000 | 318,000 | $0.9300 |
| High (complex) | 650,000 | 40,000 | 690,000 | $2.03 |
Typical cost by model
The same typical token estimate priced across a few common coding models.
| Model | Typical cost | Range (low–high) |
|---|---|---|
| GPT-4o | $0.9300 | $0.3800 – $2.03 |
| GPT-4o mini | $0.0558 | $0.0228 – $0.1215 |
| Claude Sonnet 4.6 | $1.17 | $0.4800 – $2.55 |
| Gemini 2.0 Flash | $0.0372 | $0.0152 – $0.0810 |
How we estimate this
A 2-point story is small but has some moving parts — a few agent turns across two or three files with light iteration. 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 small CRUD form wired to an existing API
- A new component plus its unit test
- A bug fix that spans two files
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 2-point user story take to build with AI?+
A 2-point user story typically takes about 318,000 tokens to build with an AI coding agent (roughly 128,000–690,000 depending on complexity), about $0.9300 at today's GPT-4o prices ($0.3800–$2.03). 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?+
A 2-point story is small but has some moving parts — a few agent turns across two or three files with light iteration. 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.