GPT-4
OpenAI · released May 28, 2023
- Type
- Proprietary
- Input
- $30.00
- Output
- $60.00
- Cached
- Not published
List price · per 1M tokens · OpenAI at 8K context · source ↗
Our take
Written Aug 3, 2026GPT-4 is OpenAI's hosted-only flagship text model from 2023, still competitive in coding tasks but limited to 8,191 tokens per request and available only through three identically priced providers. Its Arena scores have stayed remarkably stable, though its creative writing lags well behind its coding performance.
Choose this for legacy OpenAI API compatibility where short requests suffice, or for coding tasks where its Arena score in that category is strongest. Pick Azure or OpenAI direct hosting if you need known throughput at 7 or 23 tokens per second respectively. Skip it if you need longer context, multimodal input, or if cost is a primary concern — there is no cheaper tier available.
The case for it
- Coding remains its standout Arena category, with a score more than 60 points above its own creative writing result.
- Week-over-week scoring shows minimal drift — its overall Arena text score moved less than a tenth of a point between the last two measured periods.
- Fastest verified throughput on OpenAI's own API, at more than three times the speed of the Azure-hosted option.
The case against it
- Extremely expensive for a 2023-era model with limited context, with identical rates across all three providers and no cheaper tier.
- Narrow context window by current standards — 8,191 tokens with no longer variant offered.
- Creative writing is its weakest Arena category, scoring more than 60 points below its own coding result.
How good is it?
IntelligencePuzzles, maths, exam questions
Arena Text (overall)130th of 143 · 1287.1
CodingWriting and fixing code on its own
Arena Coding126th of 143 · 1329.1
Arena Coding is the only board that has scored it for this.
AgenticPlanning, calling tools, staying on task
GPT-4 is not on Arena Agent (IPS), which is where the rating would come from, so there is no rating here. It is on SWE-bench Verified, in 35th of 39 with 22.4.
WritingWe do not rate this
Two boards come close and neither tests writing: Arena Creative Writing asks people which of two replies they prefer, and LiveBench Language tests whether a model understood a passage. So we show where GPT-4 placed and give it no mark out of five.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done, which is why they get no rating.
Every published score for this model7 scoresEvery figure we hold, from 7 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Or rent it from someone else
Why this differs from the header. The strip above quotes OpenAI's own list price. This is the cheapest live offer at the widest standard context we hold, whoever is serving it — a reseller undercutting a lab is ordinary commerce, not an error.
- per 1M tokens
- $30.00 in / $60.00 out
- Context served
- 8K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouter | $30.00 / $60.00 | 8K | not measured | Unknown | Unknown | Unknown |
| Microsoft Azure AI | $30.00 / $60.00 | 8K | 18 tok/s | No | No | Confirmed |
| OpenAI | $30.00 / $60.00 | 8K4K out | 28 tok/s | No | Yesunknown period | Unknown |
Microsoft Azure AIAcross the 3 listings we hold: 2 say they do not train on prompts, 0 say they do and 1 do not say. 1 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
- ✓
- Supported
- ✗
- Not supported
- Not published
- host gave no parameter list
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouter | ✓ | ✓ | ✓ |
| Microsoft Azure AI | ✓ | ✓ | ✗ |
| OpenAI | ✓ | ✓ | ✓ |
Tool calling: 3 of 3 listings say yes. JSON output: 3 of 3 listings say yes. Strict schema: 2 of 3 listings say yes, 1 says no.
Models people weigh against GPT-4
When we formed this view
Dates behind this page
Prices last checked 9d ago
What we do not know about this model yet
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 3 listings do not say whether they train on prompts.
- We hold no cached-input rate for any of its listings.
Licence and identifiers
What the licence allowsWe hold no licence record for this model. Inside are the identifiers you need to pull it — its Hugging Face repo where we have one, our slug and a machine-readable card.
Licence
Commercial API terms. We hold no licence record for this model, so there is nothing to summarise here.
Identifiers
- Modality record
- text->text
- Catalogue slug
- openai-gpt-4