See how ChatGPT, Perplexity and Google AI Overviews describe you today
Free AI Visibility AuditGPT-4.1 is OpenAI's flagship large language model for instruction following, software engineering work and long-context reasoning, with a context window of around a million tokens and support for text, image and file inputs.
OpenAIVerified 24/09/2026Released 14/04/202501 / decision
Each figure sits next to the middle of the field, so it can be read as dear or cheap, wide or narrow, rather than floating on its own. Compared against the 126 models we track, not against an absolute standard.
Capability
Not measured
No published benchmark scores yet
Input / 1M tokens
$2.00
field median $0.30
Output / 1M tokens
$8.00
field median $1.25
Context
1,048K
field median 524K
02 / overview
Four questions, answered separately, because somebody arrives at one of them rather than at the top of the page.
The job it was built for, and the job it is not for.
GPT-4.1 is a general-purpose flagship model tuned for three jobs in particular: following detailed instructions precisely, handling real-world coding tasks, and reasoning over very long inputs. OpenAI positions it as an improvement on GPT-4o across those areas. It is not a dedicated reasoning model and not a small, cheap classifier, so high-volume routing or trivial extraction work belongs elsewhere.
Where it sits in its line, and when a sibling is the better pick.
GPT-4.1 was first seen in April 2025 as the flagship of its generation at OpenAI, replacing GPT-4o as the default choice for serious work. Pick it when the task involves a large body of source material, a long codebase or a prompt with many rules that must actually be obeyed. It is the wrong pick when the input is short and the task is simple, where a smaller sibling does the same job for less.
The API, the apps it powers, and what its limits let you do.
GPT-4.1 is reached through the OpenAI API and is available in the products and tools built on it. It accepts text, images and files in the same request, so a prompt can carry a document, a screenshot and written instructions together. The very large context window means an entire repository, a full contract set or a year of support tickets can go in whole rather than being chunked and retrieved piece by piece.
What changes because this exists, or why it does not.
The combination of a million-token window with stronger instruction following changes what you can hand a model in one go. Work that previously needed a retrieval pipeline, a chunking strategy and a reconciliation step can be done in a single call, which removes a category of engineering and a category of bug. For software teams, the same property means the model can see the surrounding code rather than guessing at it.
Superseded by GPT-4.1 Mini. See the whole line.
03 / evidence
Split by category, so a strong number never hides a thin evidence base. Verified means we read it on the benchmark's own published results; a provider's claim about its own model is shown and labelled rather than dropped.
No published benchmark scores for this model yet.
We publish a score only where we can link the result to where it was published. Until a benchmark result for this model exists in a source we read, this section stays empty rather than being filled with a provider's marketing figure.
How we decide what counts as evidence04 / ledger
Every published row, grouped by category, each compared with the best published score on the same benchmark. 'Is 64% good' is a question nobody can answer; '26 points behind the leader' is one anybody can.
Nothing in the ledger yet.
Each row here carries a score, the benchmark it came from, what the leading model scored on the same test, and a link to the published result. Rows appear as results are published and read.
How we decide what counts as evidence05 / capability
Where this model is strong, and against how many peers. Ranks are against models with evidence in that category, not against everything we track: ranking against models nobody tested would rank who published, not who is better.
No category scores to shape yet.
A category score is the weighted mean of the benchmarks published for it. With no published rows there is nothing to average, and an empty chart drawn at zero would say something false.
How we decide what counts as evidence06 / cost
List API rates as last read from the provider, with the source on every row, plus every change we have recorded since we started tracking it.
| Charge | Price | Unit | Read on | Source |
|---|---|---|---|---|
| Input | $2.00 | per 1M tokens | 2026-09-24 | Check |
| Output | $8.00 | per 1M tokens | 2026-09-24 | Check |
GPT-4.1 sits in the mid range for a flagship model, cheaper to run than the premium tier but well above the small, fast models used for bulk classification. The economics only work in its favour when you are genuinely using the long context or the instruction precision, otherwise you are paying flagship rates for work a smaller model would finish.
| Workload | Input / month | Output / month | Cost |
|---|---|---|---|
| A small product team | 20M tokens | 5M tokens | $80.00 |
| A busy support assistant | 200M tokens | 40M tokens | $720.00 |
| A document pipeline | 1000M tokens | 100M tokens | $2,800.00 |
List API rates, no caching and no batch discount, which both providers offer and which change the answer a great deal. Treat these as the ceiling, not the bill.
07 / specs
As listed by the provider's own catalogue and re-read every few hours. Anything absent is absent there too.
| Context window | 1,047,576 tokens |
|---|---|
| Maximum output | 32,768 tokens |
| Modalities | image, text, file |
| Released | 14/04/2025 |
| Status | Current |
| Catalogue identifier | openai/gpt-4.1 |
08 / lineage
What this model replaced, what replaced it, and what else its provider has in the field.
Also from OpenAI
09 / line
Every model in this family in release order, so a page from eight months ago says in one glance that two newer ones exist.
Came before
The first we track in this line.
Came after
GPT-4.1 Mini →Ordered by release date and worked out from the naming, so a new member slots in as soon as its page exists. A retired model keeps its page and its place in the line.
10 / notes
What this model changes for a brand trying to be cited in AI answers, and every change we have logged since it launched.
What it changes for you
When a model can hold a million tokens of context, the answer a buyer gets is assembled from far more source material than a short-window model could consider, so thin or fragmentary coverage of your product gets diluted by whatever else is in the window. The practical response is depth: documentation, pricing detail and comparison material that hold up when read in full rather than skimmed in a snippet. Consistency across your own pages matters more too, because a model reading all of them at once will notice where they disagree.
Buyers meet GPT-4.1 mostly through the OpenAI API, embedded inside the SaaS tools and coding assistants their teams already use, often without the model being named in the interface. It also sits behind assistant and search features built on OpenAI infrastructure, so an answer a buyer reads during research may well have been written by it.
Nothing published here yet. Changes appear within hours of a provider announcing them.
11 / questions
The things people ask about this model, answered from what is on this page rather than from anywhere else.
Models change what gets cited. We measure whether AI answers name your brand or your competitors across every assistant, and show you what to change.