GPT-5.2-Codex
OpenAI · released Jan 14, 2026
- Type
- Closed
- Input
- None held
- Output
- None held
- Cached
- None held
We don't hold a list price for this model yet · hosted only — we have no record of published weights
Our take
Written Sep 17, 2026GPT-5.2-Codex is a hosted-only coding and reasoning model: we list no download for it, so using it means choosing a host. Its measured coding and maths results are strong, but the licence terms are not disclosed in our data, so commercial use is unverified.
Use it for code generation and completion, or for mathematical and reasoning work, where its LiveBench averages are the relevant evidence. Reach it through one of the four hosted offers rather than running it yourself. Skip it if you need to run the model on your own hardware, or if you need a permissive or clearly stated licence.
The case for it
- Strong measured coding ability on set-piece generation tasks: 83.62% average over LiveBench code generation and completion, which covers writing code rather than fixing issues in an existing project.
- Strong measured mathematics ability: 88.77% average over LiveBench competition and olympiad maths tasks.
- Resolves real GitHub issues end-to-end in a scaffolded setting: 72.8% on SWE-bench Verified, measured with the mini-SWE-agent scaffold, so the figure is for the model inside that harness.
- Long documents need not be split up first, though reliable recall across all of it is unverified in our data.
The case against it
- Agentic coding is a weaker measured area: 49.39% average over LiveBench agentic coding tasks, run inside an agent harness, well below its 83.62% on code generation and completion.
- Instruction following sits at 66.45% on LiveBench constrained-rewriting tasks, below its other LiveBench category scores.
- We list no download for it, so using it means choosing a host, and the licence is not disclosed in our data, so commercial-use terms are unverified.
How good is it?
EverydayGeneral questions and everyday reasoning
Not yet scored on Arena Text (overall). It is on LiveBench Data Analysis, in 18th of 58 with 78.2.
CodingWriting and fixing code on its own
Not yet scored on Arena Coding. It is on Arena Code (WebDev), in 78th of 95 with 1339.
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent. It is on LiveBench Agentic Coding, in 37th of 58 with 49.39.
WritingDrafting and rewriting prose
Not yet scored on Arena Creative Writing. It is on LiveBench Language, in 47th of 58 with 73.68.
Boards this model appears on that none of the ratings above are built on.
Every published score for this model10 scoresEvery figure we hold, from 10 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Where to rent it
Prices checked 4 hours ago — each listing carries its own date.
- per 1M tokens
- $1.75 in / $14.00 out
- Context served
- 400K
- Throughput
- ~33 tok/s
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouterOpenRouter's own listing | $1.75 / $14.00checked 4 hours ago | 400K | not measured | Unknown | Unknown | Unknown |
| Microsoft Azure AIThrough OpenRouter | $1.75 / $14.00checked 4 hours ago | 400K128K max reply | 33 tok/s | No | No | Confirmed |
OpenRouterAcross the 2 listings we hold: 1 says it does not train on prompts, 0 say they do and 1 does not say. 1 appears in the zero-retention registry we check; the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| Microsoft Azure AIThrough OpenRouter | ✓ | ✓ | ✓ |
Tool calling: 2 of 2 listings say yes. JSON output: 2 of 2 listings say yes. Strict schema: 2 of 2 listings say yes.
Models people weigh against GPT-5.2-Codex
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
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 2 listings does not say whether it trains on prompts.
- We don't hold a list price for this model yet — the gap is ours, not the lab's.
- We hold no batch or off-peak 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
We hold no licence record for this model, and no record of published weights either — so we can neither summarise its terms nor point you at the weights.
Identifiers
- Takes in, gives back
- Text and images in, text out
- Catalogue slug
- openai-gpt-5-2-codex