OpenAIGPT-5.4 and GPT-5.4 Pro

The GPT-5 learning curve

GPT-5 rewards precise scope, consistent instructions, and deliberate use of reasoning and verbosity controls.

Start with the promptReview the evidence

The mental model

What changes the result

“Make the task unambiguous, then tune effort with evals instead of adding prompt folklore.”

Three levels

Climb the curve in this order

Do not add production scaffolding before you can describe and measure a good single-task result.

  1. 01

    Write it like a good issue

    Give the model enough context to act without making it infer the job.

    • Name the outcome, relevant inputs, constraints, and output format.
    • Point to an existing example when you want the same pattern.
    • Remove conflicting instructions before trying stronger wording.
    Ready to move on when: The request has one clear interpretation and one observable completion state.
  2. 02

    Tune the controls

    Match intelligence, latency, and response length to the work.

    • Use lower reasoning effort for routine transformations and higher effort for complex multistep work.
    • Control visible answer length with verbosity or an explicit format instruction.
    • Build a small eval set of real examples before changing a production prompt.
    Ready to move on when: You can explain why the chosen effort and verbosity beat a cheaper baseline.
  3. 03

    Design an agent loop

    Preserve useful reasoning context and make tool use observable.

    • Use the Responses API for stateful, tool-using workflows.
    • Define tool boundaries and completion conditions instead of asking the model to “do everything.”
    • Inspect failed traces, simplify the prompt, and lock successful behavior into templates.
    Ready to move on when: Failures are reproducible, categorized, and covered by evals before the next prompt revision.

Starter template

Begin here, then earn every extra instruction

Replace the bracketed text, run it on a real task, and add guidance only when you can name the failure it fixes.

gpt-5-starter.txt
<task>
[One concrete outcome]
</task>

<context>
[Only the facts, files, or examples needed for this task]
</context>

<constraints>
- [Important boundary]
- [Required source or tool behavior]
</constraints>

<deliverable>
[Format, audience, length, and acceptance criteria]
</deliverable>

Patterns that work

Issue-shaped requests

File paths, components, examples, acceptance checks, and a bounded outcome give coding work useful structure.

Metaprompting with evidence

Ask the model to diagnose failed outputs and improve the prompt, then verify the revision against an eval set.

Separate effort from verbosity

A task can require deep reasoning and still need a short final answer; control those dimensions independently.

Common failure modes

Conflicting rules

The model spends effort reconciling incompatible tone, scope, or tool instructions.

Fix: Consolidate instructions and state which rule wins when an exception is intentional.

Maximum effort by default

Routine work becomes slower and more expensive without a measurable quality gain.

Fix: Start with the lowest plausible effort and raise it only when evals show under-thinking.

One heroic prompt

A large prompt becomes brittle across unrelated tasks and hard to debug.

Fix: Use stable shared context plus smaller task templates with their own success criteria.

Current as of July 26, 2026

What the latest guidance changes

OpenAI’s guidance for the GPT-5 series emphasizes baseline evals, clearer and often shorter prompts, reusable templates, and independent controls for reasoning effort and visible verbosity.

Evidence ledger

Read the sources behind the advice

Official provider documentation takes priority. Maintainer field notes are useful evidence, but they are labeled separately.

GPT-5.4 model reference

official

OpenAI

Current flagship model identity and supported API capabilities.

A practical guide to building with GPT-5

official

OpenAI

Evals, prompt simplification, metaprompting, effort, verbosity, and Responses API guidance.

How OpenAI uses Codex

official

OpenAI

Issue-shaped tasks, AGENTS.md context, environment improvement, and iterative workflows.

This guide is reviewed at least every 30 days. Automated checks flag unavailable sources and overdue reviews; substantive advice changes require human verification.