GPT-5.3-Codex
OpenAI · released Feb 24, 2026
- Type
- Proprietary
- Input
- $1.75
- Output
- $14.00
- Cached
- $0.17
List price · per 1M tokens · OpenAI at 400K context · source ↗
Our take
Written Aug 3, 2026GPT-5.3-Codex is a proprietary coding model from OpenAI that can handle up to 400,000 tokens in a single request and accepts text, images and files. It is built for large-codebase work and web development, with a premium throughput tier available from the same provider.
Pick this when you are working with substantial code repositories that need long-context review, or when you want throughput optionality from one provider without switching hosts. Use it for web development tasks where its stable Arena Code scores are relevant. Skip it if you need open weights for self-hosting, if your budget is sensitive to high output costs, or if you need general chat or reasoning benchmarks rather than coding-specific ones.
The case for it
- Extremely long context for code review and large repositories, at 400,000 tokens.
- Stable leaderboard position in web development coding: Arena Code Elo held within 0.08 points across six evaluations over one week.
- Throughput optionality on the same provider, with a tier roughly one-and-a-half times faster than the standard one.
The case against it
- Output costs eight times the input rate at every tier, making it expensive for generation-heavy workloads.
- The premium tier doubles the price while increasing throughput by a smaller factor, so the speed-per-dollar ratio worsens.
- Arena Code Elo flat at about 1409 with no upward trend across six daily evaluations.
How good is it?
IntelligencePuzzles, maths, exam questions
Nobody we watch has scored GPT-5.3-Codex for this. We would take the rating from Arena Text (overall).
CodingWriting and fixing code on its own
GPT-5.3-Codex is not on Arena Coding, which is where the rating would come from, so there is no rating here. It is on Arena Code (WebDev), in 38th of 74 with 1408.9.
AgenticPlanning, calling tools, staying on task
Nobody we watch has scored GPT-5.3-Codex for this. We would take the rating from Arena Agent (IPS).
WritingWe do not rate this
Nobody we watch has scored this model for writing. 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.
Every published score for this model1 scoreEvery figure we hold, from 1 board, 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
- $1.75 in / $14.00 out
- Context served
- 400K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouter | $1.75 / $14.00 | 400K | not measured | Unknown | Unknown | Unknown |
| Microsoft Azure AI | $1.75 / $14.00 | 400K | 57 tok/s | No | No | Confirmed |
| OpenAI | $1.75 / $14.00 | 400K | 70 tok/s | No | Yesunknown period | Unknown |
| OpenAIpriority tier | $3.50 / $28.00 | 400K | 99 tok/s | No | Yesunknown period | Unknown |
Microsoft Azure AIAcross the 4 listings we hold: 3 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 | ✓ | ✓ | ✓ |
| OpenAIpriority | ✓ | ✓ | ✓ |
Tool calling: 4 of 4 listings say yes. JSON output: 4 of 4 listings say yes. Strict schema: 4 of 4 listings say yes.
Models people weigh against GPT-5.3-Codex
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 4 listings do not say whether they train on prompts.
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+image+file->text
- Catalogue slug
- openai-gpt-5-3-codex