Documentation — Spybara

deprecations.md +86 −72

Details

135To improve reliability and make it easier for developers to choose the right models, we are deprecating a set of older OpenAI models. Access to these models will be shut down on the dates below.135To improve reliability and make it easier for developers to choose the right models, we are deprecating a set of older OpenAI models. Access to these models will be shut down on the dates below.

136 136

138| ------------- | ---------------------------------------------------------------------- | ------------------------------------- |138| ---------------- | ---------------------------------------------------------------------- | ------------------------------------- |

141| October 23, 2026 | `gpt-4-1106-preview` | `gpt-5.6-sol` |

144| October 23, 2026 | `gpt-4o-2024-05-13` | `gpt-5.6-sol` |

145| October 23, 2026 | `gpt-image-1` | `gpt-image-2` |

146| October 23, 2026 | `o1-2024-12-17` \| `o1` | `gpt-5.6-sol` |

149| October 23, 2026 | `ft-o4-mini-2025-04-16` | `gpt-5.6-terra` |

151

152We are also removing fine-tuned versions as below:

153

154| Shutdown date | Model snapshot | Recommended replacement base model |

155| ---------------- | ---------------------------- | ---------------------------------- |

156| October 23, 2026 | `ft-gpt-3.5-turbo` | `gpt-5.4-mini` |

157| October 23, 2026 | `ft-gpt-4` | `gpt-5.5` |

158| October 23, 2026 | `ft-gpt-4.1-nano-2025-04-14` | `gpt-5.4-nano` |

159| October 23, 2026 | `ft-babbage-002` | `gpt-5.4-mini` |

160| October 23, 2026 | `ft-davinci-002` | `gpt-5.4-mini` |

161

162### 2026-03-24: Sora 2 video generation models and Videos API

163

164On March 24th, 2026, we notified developers using the Videos API and Sora 2 video generation model aliases and snapshots of their deprecation and removal from the API on September 24, 2026.

165

166| Shutdown date | Model / system | Recommended replacement |

167| ------------- | ----------------------- | ----------------------- |

168| 2026-09-24 | Videos API | --- |

169| 2026-09-24 | `sora-2` | --- |

170| 2026-09-24 | `sora-2-pro` | --- |

171| 2026-09-24 | `sora-2-2025-10-06` | --- |

172| 2026-09-24 | `sora-2-2025-12-08` | --- |

173| 2026-09-24 | `sora-2-pro-2025-10-06` | --- |

174

175### 2025-09-26: Legacy GPT model snapshots

176

177To improve reliability and make it easier for developers to choose the right models, we are deprecating a set of older OpenAI models with declining usage over the next six to twelve months. Access to these models will be shut down on the dates below.

178

179| Shutdown date | Model / system | Recommended replacement |

180| ------------- | ------------------------ | ------------------------------ |

181| 2026-09-28 | `gpt-3.5-turbo-instruct` | `gpt-5.4-mini` or `gpt-5-mini` |

182| 2026-09-28 | `babbage-002` | `gpt-5.4-mini` or `gpt-5-mini` |

183| 2026-09-28 | `davinci-002` | `gpt-5.4-mini` or `gpt-5-mini` |

184| 2026-09-28 | `gpt-3.5-turbo-1106` | `gpt-5.4-mini` or `gpt-5-mini` |

185

186### 2025-08-20: Assistants API

187

188On August 26th, 2025, we notified developers using the Assistants API of its deprecation and removal from the API one year later, on August 26, 2026.

189

190When we released the [Responses API](https://developers.openai.com/api/reference/resources/responses/methods/create) in [March 2025](https://developers.openai.com/api/docs/changelog), we announced plans to bring all Assistants API features to the easier to use Responses API, with a sunset date in 2026.

191

192See the Assistants to Conversations [migration guide](https://developers.openai.com/api/docs/assistants/migration) to learn more about how to migrate your current integration to the Responses API and Conversations API.

193

194| Shutdown date | Model / system | Recommended replacement |

195| ------------- | -------------- | ----------------------------------- |

196| 2026‑08‑26 | Assistants API | Responses API and Conversations API |

197

198## Past deprecations

199

200Past deprecations are listed below, with the most recent announcements at the top.

201

202### 2026-04-22: Legacy GPT model snapshots (July 2026 shutdown)

203

204On April 22, 2026, we announced the deprecation of the following older OpenAI models. Access to these models was shut down on July 23, 2026.

205

206| Shutdown date | Model snapshot | Substitute model |

207| ------------- | ------------------------------------------------------------- | ---------------------------- |

155| 2026-10-23 | `gpt-4-0613` \| `gpt-4`, `gpt-4-0613-completions`, `gpt-4-completions` | `gpt-5.6-sol` |

156| 2026-10-23 | `gpt-4-1106-preview` | `gpt-5.6-sol` |

159| 2026-10-23 | `gpt-4o-2024-05-13` | `gpt-5.6-sol` |

160| 2026-10-23 | `gpt-image-1` | `gpt-image-2` |

161| 2026-10-23 | `o1-2024-12-17` \| `o1` | `gpt-5.6-sol` |

164| 2026-10-23 | `ft-o4-mini-2025-04-16` | `gpt-5.6-terra` |

166 223

167We are also removing fine-tuned versions as below:224### 2025-11-18: chatgpt-4o-latest snapshot

168 225

169| Shutdown date | Model snapshot | Recommended replacement base model |226On November 18th, 2025, we notified developers using `chatgpt-4o-latest` model snapshot of its deprecation and removal from the API on February 17, 2026.

170| ------------- | ---------------------------- | ---------------------------------- |

171| 2026-10-23 | `ft-gpt-3.5-turbo` | `gpt-5.4-mini` |

172| 2026-10-23 | `ft-gpt-4` | `gpt-5.5` |

173| 2026-10-23 | `ft-gpt-4.1-nano-2025-04-14` | `gpt-5.4-nano` |

174| 2026-10-23 | `ft-babbage-002` | `gpt-5.4-mini` |

175| 2026-10-23 | `ft-davinci-002` | `gpt-5.4-mini` |

176 227

177### 2026-03-24: Sora 2 video generation models and Videos API228| Shutdown date | Model / system | Recommended replacement |

229| ------------- | ------------------- | ----------------------- |

230| 2026-02-17 | `chatgpt-4o-latest` | `gpt-5.1-chat-latest` |

178 231

179On March 24th, 2026, we notified developers using the Videos API and Sora 2 video generation model aliases and snapshots of their deprecation and removal from the API on September 24, 2026.232### 2025-11-17: codex-mini-latest model snapshot

233

234On November 17th, 2025, we notified developers using `codex-mini-latest` model of its deprecation and removal from the API on February 12, 2026. As part of this deprecation, we will no longer support our legacy local shell tool, which is only available for use with `codex-mini-latest`. For new use cases, please use our latest shell tool.

180 235

182| ------------- | ----------------------- | ----------------------- |237| ------------- | ------------------- | ----------------------- |

183| 2026-09-24 | Videos API | --- |238| 2026-02-12 | `codex-mini-latest` | `gpt-5-codex-mini` |

184| 2026-09-24 | `sora-2` | --- |

185| 2026-09-24 | `sora-2-pro` | --- |

186| 2026-09-24 | `sora-2-2025-10-06` | --- |

187| 2026-09-24 | `sora-2-2025-12-08` | --- |

188| 2026-09-24 | `sora-2-pro-2025-10-06` | --- |

189 239

190### 2025-11-14: DALL·E model snapshots240### 2025-11-14: DALL·E model snapshots

191 241

196| 2026-05-12 | `dall-e-2` | `gpt-image-2`, `gpt-image-1`, or `gpt-image-1-mini` |246| 2026-05-12 | `dall-e-2` | `gpt-image-2`, `gpt-image-1`, or `gpt-image-1-mini` |

197| 2026-05-12 | `dall-e-3` | `gpt-image-2`, `gpt-image-1`, or `gpt-image-1-mini` |247| 2026-05-12 | `dall-e-3` | `gpt-image-2`, `gpt-image-1`, or `gpt-image-1-mini` |

198 248

199### 2025-09-26: Legacy GPT model snapshots249### 2025-09-26: Legacy GPT model snapshots (March 2026 shutdown)

200 250

201To improve reliability and make it easier for developers to choose the right models, we are deprecating a set of older OpenAI models with declining usage over the next six to twelve months. Access to these models will be shut down on the dates below.251To improve reliability and make it easier for developers to choose the right models, we deprecated a set of older OpenAI models with declining usage. Access to these models was shut down on March 26, 2026.

202 252

204| ------------- | -------------------------------------------------------------------------------------------------------------------------- | ------------------------------ |254| ------------- | -------------------------------------------------------------------------------------------------------------------------- | ----------------------- |

205| 2026‑03‑26 | `gpt-4-0314` | `gpt-5` or `gpt-4.1*` |255| 2026‑03‑26 | `gpt-4-0314` | `gpt-5` or `gpt-4.1*` |

206| 2026‑03‑26 | `gpt-4-1106-preview` | `gpt-5` or `gpt-4.1*` |256| 2026‑03‑26 | `gpt-4-1106-preview` | `gpt-5` or `gpt-4.1*` |

207| 2026‑03‑26 | `gpt-4-0125-preview` (including `gpt-4-turbo-preview` and `gpt-4-turbo-preview-completions`, which point to this snapshot) | `gpt-5` or `gpt-4.1*` |257| 2026‑03‑26 | `gpt-4-0125-preview` (including `gpt-4-turbo-preview` and `gpt-4-turbo-preview-completions`, which point to this snapshot) | `gpt-5` or `gpt-4.1*` |

208| 2026-09-28 | `gpt-3.5-turbo-instruct` | `gpt-5.4-mini` or `gpt-5-mini` |

209| 2026-09-28 | `babbage-002` | `gpt-5.4-mini` or `gpt-5-mini` |

210| 2026-09-28 | `davinci-002` | `gpt-5.4-mini` or `gpt-5-mini` |

211| 2026-09-28 | `gpt-3.5-turbo-1106` | `gpt-5.4-mini` or `gpt-5-mini` |

212 258

213\*For tasks that are especially latency sensitive and don't require reasoning259\*For tasks that are especially latency sensitive and don't require reasoning

214 260

222| ------------- | ------------------------ | ----------------------- |268| ------------- | ------------------------ | ----------------------- |

223| 2026‑05‑12 | OpenAI-Beta: realtime=v1 | Realtime API |269| 2026‑05‑12 | OpenAI-Beta: realtime=v1 | Realtime API |

224 270

225### 2025-08-20: Assistants API

~~226~~

227On August 26th, 2025, we notified developers using the Assistants API of its deprecation and removal from the API one year later, on August 26, 2026.

~~228~~

229When we released the [Responses API](https://developers.openai.com/api/reference/resources/responses/methods/create) in [March 2025](https://developers.openai.com/api/docs/changelog), we announced plans to bring all Assistants API features to the easier to use Responses API, with a sunset date in 2026.

~~230~~

231See the Assistants to Conversations [migration guide](https://developers.openai.com/api/docs/assistants/migration) to learn more about how to migrate your current integration to the Responses API and Conversations API.

~~232~~

233| Shutdown date | Model / system | Recommended replacement |

234| ------------- | -------------- | ----------------------------------- |

235| 2026‑08‑26 | Assistants API | Responses API and Conversations API |

~~236~~

237### 2025-09-15: gpt-4o-realtime-preview models271### 2025-09-15: gpt-4o-realtime-preview models

238 272

239In September, 2025, we notified developers using gpt-4o-realtime-preview models of their deprecation and removal from the API in six months.273In September, 2025, we notified developers using gpt-4o-realtime-preview models of their deprecation and removal from the API in six months.

247| 2026-05-07 | gpt-4o-audio-preview | gpt-audio-1.5 |281| 2026-05-07 | gpt-4o-audio-preview | gpt-audio-1.5 |

248| 2026-05-07 | gpt-4o-mini-audio-preview | gpt-audio-mini |282| 2026-05-07 | gpt-4o-mini-audio-preview | gpt-audio-mini |

249 283

250## Past deprecations

~~251~~

252Past deprecations are listed below, with the most recent announcements at the top.

~~253~~

254### 2025-11-18: chatgpt-4o-latest snapshot

~~255~~

256On November 18th, 2025, we notified developers using `chatgpt-4o-latest` model snapshot of its deprecation and removal from the API on February 17, 2026.

~~257~~

258| Shutdown date | Model / system | Recommended replacement |

259| ------------- | ------------------- | ----------------------- |

260| 2026-02-17 | `chatgpt-4o-latest` | `gpt-5.1-chat-latest` |

~~261~~

262### 2025-11-17: codex-mini-latest model snapshot

~~263~~

264On November 17th, 2025, we notified developers using `codex-mini-latest` model of its deprecation and removal from the API on February 12, 2026. As part of this deprecation, we will no longer support our legacy local shell tool, which is only available for use with `codex-mini-latest`. For new use cases, please use our latest shell tool.

~~265~~

266| Shutdown date | Model / system | Recommended replacement |

267| ------------- | ------------------- | ----------------------- |

268| 2026-02-12 | `codex-mini-latest` | `gpt-5-codex-mini` |

~~269~~

270### 2025-06-10: gpt-4o-realtime-preview-2024-10-01284### 2025-06-10: gpt-4o-realtime-preview-2024-10-01

271 285

272On June 10th, 2025, we notified developers using gpt-4o-realtime-preview-2024-10-01 of its deprecation and removal from the API in three months.286On June 10th, 2025, we notified developers using gpt-4o-realtime-preview-2024-10-01 of its deprecation and removal from the API in three months.

464 478

465### Plain-text aliases479### Plain-text aliases

466 480

467- computer-use-preview-2025-03-11 | computer-use-preview

468- o3-deep-research-2025-06-26 | o3-deep-research

469- o4-mini-deep-research-2025-06-26 | o4-mini-deep-research

470- gpt-3.5-turbo-0125 | gpt-3.5-turbo, gpt-3.5-turbo-completions481- gpt-3.5-turbo-0125 | gpt-3.5-turbo, gpt-3.5-turbo-completions

471- gpt-4-0613 | gpt-4, gpt-4-0613-completions, gpt-4-completions482- gpt-4-0613 | gpt-4, gpt-4-0613-completions, gpt-4-completions

472- gpt-4-turbo | gpt-4-turbo-2024-04-09, gpt-4-turbo-completions483- gpt-4-turbo | gpt-4-turbo-2024-04-09, gpt-4-turbo-completions

475- o1-pro-2025-03-19 | o1-pro486- o1-pro-2025-03-19 | o1-pro

476- o3-mini-2025-01-31 | o3-mini487- o3-mini-2025-01-31 | o3-mini

477- o4-mini-2025-04-16 | o4-mini488- o4-mini-2025-04-16 | o4-mini

489- computer-use-preview-2025-03-11 | computer-use-preview

490- o3-deep-research-2025-06-26 | o3-deep-research

491- o4-mini-deep-research-2025-06-26 | o4-mini-deep-research

guides/content-provenance.md +278 −0 created

Details

1# Content provenance

3> For the complete documentation index, see [llms.txt](/llms.txt). Markdown versions of documentation pages are available by appending `.md` to the page URL.

5Use the Content Provenance API to check whether an image or audio file contains

6supported OpenAI provenance signals. Send a file to

7`POST /v1/content_provenance_checks` to receive the completed verification

8results in the same response. Use these signals in content review,

9fact-checking, labeling, and trust and safety workflows.

11To check a file in your browser, use the web tool at

12[openai.com/verify](https://openai.com/verify/).

14For request parameters and response schemas, see the

15[Content provenance API reference](https://developers.openai.com/api/reference/resources/content_provenance_checks/methods/create).

17A `not_detected` result means the tool didn't find supported signals in the

18 uploaded file. Content may still have been generated by OpenAI if its metadata

19 was stripped or shows evidence of tampering, its watermark was degraded, it

20 came from a legacy generation model, or it was created before provenance

21 signals were available. The tool doesn't currently detect content generated by

22 another company's AI model, so a `not_detected` result doesn't rule that out

23 either.

25## What content provenance checks

27Content provenance checks supported files for the following signals:

29| Signal | Applies to | What it checks |

30| ------------------------ | ---------------- | ------------------------------------------------ |

31| C2PA Content Credentials | Images | Signed metadata with issuer and AI-use details |

32| SynthID | Images and audio | A watermark embedded directly in supported media |

34C2PA metadata provides more context about a file's origin. Editing, converting,

35or sharing a file can remove its metadata. A SynthID watermark is part of the

36image or audio itself and may survive some transformations.

38The API checks for supported OpenAI signals. It isn't a general-purpose AI

39detector and doesn't identify content generated by every AI system. Visible

40watermarks and labels are separate from the provenance signals checked by the

41API.

43## Verify a file

45Send an image or audio file as the `file` field with the OpenAI SDK. The SDK

46builds the multipart request and reads your API key from the `OPENAI_API_KEY`

47environment variable:

49Verify an image

51```python

52from openai import OpenAI

54client = OpenAI()

56with open("./example.png", "rb") as image:

57 result = client.content_provenance_checks.create(

58 file=("example.png", image, "image/png"),

59 )

61print(result)

62```

64```go

65package main

67import (

68 "context"

69 "fmt"

70 "os"

72 "github.com/openai/openai-go/v3"

73)

75func main() {

76 client := openai.NewClient()

78 image, err := os.Open("./example.png")

79 if err != nil {

80 panic(err)

81 }

82 defer image.Close()

84 result, err := client.ContentProvenanceChecks.New(

85 context.Background(),

86 openai.ContentProvenanceCheckNewParams{

87 File: openai.File(image, "example.png", "image/png"),

88 },

89 )

90 if err != nil {

91 panic(err)

92 }

94 fmt.Println(result)

95}

96```

98```ruby

99require "openai"

100require "pathname"

101

102client = OpenAI::Client.new

103image = OpenAI::FilePart.new(Pathname("./example.png"), content_type: "image/png")

104result = client.content_provenance_checks.create(file: image)

105

106puts result

107```

108

109```bash

110curl https://api.openai.com/v1/content_provenance_checks \

111 -H "Authorization: Bearer $OPENAI_API_KEY" \

112 -F "file=@./example.png;type=image/png"

113```

114

115

116Use these OpenAI SDK versions or later: Python 2.52.0, Go 3.49.0, and Ruby

1170.75.0.

118

119To verify Opus audio, use the same endpoint and set the uploaded file's media

120type to `audio/ogg`:

121

122```bash

123curl https://api.openai.com/v1/content_provenance_checks \

124 -H "Authorization: Bearer $OPENAI_API_KEY" \

125 -F "file=@./example.opus;type=audio/ogg"

126```

127

128The response contains the completed result. For example, an image returns:

129

130```json

131{

132 "object": "content_provenance_check",

133 "created_at": 1778000000,

134 "results": [

135 {

136 "type": "c2pa",

137 "outcome": "detected",

138 "validation_state": "trusted",

139 "issuer": "OpenAI OpCo, LLC",

140 "model": "gpt-image",

141 "generated_at": "2026-07-27T18:34:12Z"

142 },

143 {

144 "type": "synthid",

145 "outcome": "not_detected",

146 "model": null,

147 "generated_at": null

148 }

149 ]

150}

151```

152

153The `object` field identifies the response, and `created_at` is the check's

154creation time as a Unix timestamp in seconds. The entries in `results` depend on

155the uploaded file: images include C2PA and SynthID results, and audio includes a

156SynthID result. The API omits checks that don't apply instead of returning

157`not_detected`.

158

159The API completes verification before it returns. You don't need to create a

160background job, poll another endpoint, or upload the file to the Files API.

161

162If a request fails, check the HTTP status and `error.code` when available. A

163malformed, unsupported, or blocked file returns `400`; an organization without

164access receives `404`; and requests over the rate limit return `429`. Retry only

165transient failures, such as rate limits or server errors. For general guidance,

166see [API error codes](https://developers.openai.com/api/docs/guides/error-codes).

167

168## Understand verification results

169

170Read each applicable entry in `results` independently. Image results include

171C2PA and SynthID entries, while audio results include a SynthID entry. The

172response doesn't include a top-level `outcome`.

173

174### C2PA results

175

176A C2PA result describes the state of an image's Content Credentials:

177

178```json

179{

180 "type": "c2pa",

181 "outcome": "detected",

182 "validation_state": "trusted",

183 "issuer": "OpenAI OpCo, LLC",

184 "model": "gpt-image",

185 "generated_at": "2026-07-27T18:34:12Z"

186}

187```

188

189Use the fields as follows:

190

191- `outcome` indicates whether OpenAI-issued AI-generation credentials were

192 `detected` or `not_detected`.

193- `validation_state` indicates whether the manifest is `trusted`, `valid`,

194 `invalid`, or `not_present`.

195- `issuer` identifies the manifest issuer when that information is available.

196- `model` identifies the generating model when that information is available.

197- `generated_at` identifies the content's generation time when that information

198 is available.

199

200The outcome is `detected` only when a `trusted` or `valid` manifest identifies

201OpenAI as its issuer and includes an AI-generation action. A third-party

202manifest, a manifest without an AI-generation action, an `invalid` manifest, or

203a `not_present` manifest produces `not_detected`. The `issuer` and

204`validation_state` can still describe a manifest even when the outcome is

205`not_detected`.

206

207Don't treat an `invalid` manifest as reliable provenance evidence. A

208`not_present` result means the image has no available C2PA manifest.

209

210### SynthID results

211

212A SynthID result describes whether the verifier detected a supported watermark

213in an image or audio file:

214

215```json

216{

217 "type": "synthid",

218 "outcome": "detected",

219 "model": null,

220 "generated_at": null

221}

222```

223

224An outcome of `detected` means the file contains a recognized watermark. An

225outcome of `not_detected` means the verifier didn't detect that watermark. It

226doesn't rule out AI-generated or AI-modified content. `model` and

227`generated_at` provide the generating model and generation time when available;

228either field can be `null`.

229

230## Supported formats and availability

231

232The API supports the following file formats:

233

234- **Images:** PNG, JPEG, and WebP.

235- **Audio:** MP3, Opus, AAC, FLAC, WAV, and PCM.

236

237Limit each uploaded file to 50 MiB. Audio must be 60 seconds or shorter after

238decoding.

239

240Set the uploaded `file` part's media type. For example, use `image/png` for a PNG

241image or `audio/ogg` for Opus audio. Don't add a separate `type` field or

242manually set the `multipart/form-data` request header. The `curl` `-F` option

243sets the request content type and multipart boundary. Send one file per request.

244

245Content provenance checks aren't eligible for

246[Zero Data Retention](https://developers.openai.com/api/docs/guides/your-data#zero-data-retention).

247

248Strict rate limits help protect the API against misuse. Organizations can

249[apply for higher limits](https://openai.com/form/content-provenance-api/), and

250OpenAI reviews each application on a case-by-case basis.

251

252If the API returns `429 rate_limit_exceeded`, reduce your request rate and

253honor the `Retry-After` header when present. See

254[rate limits](https://developers.openai.com/api/docs/guides/rate-limits) for general retry guidance.

255

256## Use verification results responsibly

257

258Use verification results as evidence in a broader review process:

259

260- Treat `detected` as evidence of a specific supported signal, not a complete

261 history of a file.

262- Treat `not_detected` as an absence of detected evidence, not proof that the

263 content is human-created or wasn't generated with OpenAI.

264- Check the C2PA issuer before attributing an image to a particular provider.

265- Verify the original file when possible. Compression, cropping, screenshots,

266 metadata removal, and format conversions can erase or weaken a signal.

267- Account for the originating product, model, file format, and creation date.

268 Not all OpenAI-generated content contains a supported signal.

269- Pair automated decisions with human review in high-stakes workflows.

270- Don't use repeated queries to reverse-engineer, remove, or evade a watermark.

271- Don't infer a prompt, account, or individual creator from a verification

272 result.

273

274Using the Content Provenance API is subject to the

275[OpenAI Services Agreement](https://openai.com/policies/services-agreement/).

276

277For information about platform-wide monitoring and retention settings, see

278[data controls](https://developers.openai.com/api/docs/guides/your-data).

guides/fast-mode.md +1 −1

Details

36const response = await openai.responses.create({36const response = await openai.responses.create({

37 model: "gpt-5.6-sol",37 model: "gpt-5.6-sol",

38 input: "What does 'fit check for my napalm era' mean?",38 input: "What does 'fit check for my napalm era' mean?",

~~39 service_tier: "priority",~~39 service_tier: "fast",

40});40});

41 41

42console.log(response);42console.log(response);

guides/prompt-caching.md +6 −3

Details

221 221

222## Requirements222## Requirements

223 223

224Caching is available for prompts containing 1024 tokens or more.224The minimum prefix length required for caching depends on the model:

225 225

226All requests, including those with fewer than 1024 tokens, display a `cached_tokens` field in the usage token details. Responses API returns this field in `usage.input_tokens_details` on the [Response object](https://developers.openai.com/api/reference/resources/responses); Chat Completions API returns it in `usage.prompt_tokens_details` on the [Chat object](https://developers.openai.com/api/reference/resources/chat). The field indicates how many input tokens were read from cache. For requests under 1024 tokens, `cached_tokens` is zero.226- **GPT-5.6 and later models:** Caching is available for prefixes containing at least 1,024 tokens. This is a strict minimum.

227- **GPT-5.5 and earlier models:** The minimum cacheable prefix length varies by model and can range from 1,024 to 2,048 tokens. Prompts just above 1,024 tokens may not be cached consistently.

228

229All requests, including those with fewer than 1,024 tokens, display a `cached_tokens` field in the usage token details. Responses API returns this field in `usage.input_tokens_details` on the [Response object](https://developers.openai.com/api/reference/resources/responses); Chat Completions API returns it in `usage.prompt_tokens_details` on the [Chat object](https://developers.openai.com/api/reference/resources/chat). The field indicates how many input tokens were read from cache. For requests under 1,024 tokens, `cached_tokens` is zero.

227 230

228For GPT-5.6 models and later model families, `cache_write_tokens` reports the number of prompt tokens written to cache. Cache write billing uses this value at 1.25× the uncached input token rate.231For GPT-5.6 models and later model families, `cache_write_tokens` reports the number of prompt tokens written to cache. Cache write billing uses this value at 1.25× the uncached input token rate.

229 232

250 253

251- **Messages:** The complete messages array, encompassing system, user, and assistant interactions.254- **Messages:** The complete messages array, encompassing system, user, and assistant interactions.

252- **Images:** Images included in user messages, either as links or as base64-encoded data, as well as multiple images can be sent. Ensure the detail parameter is set identically, as it impacts image tokenization.255- **Images:** Images included in user messages, either as links or as base64-encoded data, as well as multiple images can be sent. Ensure the detail parameter is set identically, as it impacts image tokenization.

253- **Tool use:** Both the messages array and the list of available `tools` can be cached, contributing to the minimum 1024 token requirement.256- **Tool use:** Both the messages array and the list of available `tools` can be cached, contributing to the model's minimum cacheable prefix length.

254- **Structured outputs:** The structured output schema serves as a prefix to the system message and can be cached.257- **Structured outputs:** The structured output schema serves as a prefix to the system message and can be cached.

255 258

256## Best practices259## Best practices

guides/your-data.md +9 −11

Details

173The complete, unfiltered regional support table follows. Model snapshots for each service are listed in **API Endpoint, tool and model support**. When regional processing supports only a subset of snapshots, that subset is included in the processing-services cell.173The complete, unfiltered regional support table follows. Model snapshots for each service are listed in **API Endpoint, tool and model support**. When regional processing supports only a subset of snapshots, that subset is included in the processing-services cell.

174 174

176| -------------------------- | ------------------- | :--------------: | :-----------------: | :-----------------: | --------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |176| -------------------------- | ------------------- | :--------------: | :-----------------: | :-----------------: | --------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |

177| United States | `us.api.openai.com` | Yes | Yes | No | Text, Audio, Voice, Image | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/evals` `/v1/files` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/realtime` `/v1/realtime/transcription_sessions` `/v1/realtime/translations` `/v1/realtime (legacy previews)` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `/v1/vector_stores` `Code Interpreter tool` `File Search` `File Uploads` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/evals` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/realtime` `/v1/realtime/transcription_sessions` `/v1/realtime/translations` `/v1/realtime (legacy previews)` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `Code Interpreter tool` `File Search` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` |177| United States | `us.api.openai.com` | Yes | Yes | No | Text, Audio, Voice, Image | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/evals` `/v1/files` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/realtime` `/v1/realtime/transcription_sessions` `/v1/realtime/translations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `/v1/vector_stores` `Code Interpreter tool` `File Search` `File Uploads` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/evals` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/realtime` `/v1/realtime/transcription_sessions` `/v1/realtime/translations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `Code Interpreter tool` `File Search` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` |

178| Europe (EEA + Switzerland) | `eu.api.openai.com` | Yes | Yes | Yes\*\* | Text, Audio, Voice, Image\* | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/evals` `/v1/files` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/realtime` `/v1/realtime/transcription_sessions` `/v1/realtime/translations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `/v1/vector_stores` `Code Interpreter tool` `File Search` `File Uploads` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/evals` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/realtime` `/v1/realtime/transcription_sessions` `/v1/realtime/translations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `Code Interpreter tool` `File Search` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` |178| Europe (EEA + Switzerland) | `eu.api.openai.com` | Yes | Yes | Yes\*\* | Text, Audio, Voice, Image\* | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/evals` `/v1/files` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/realtime` `/v1/realtime/transcription_sessions` `/v1/realtime/translations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `/v1/vector_stores` `Code Interpreter tool` `File Search` `File Uploads` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/evals` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/realtime` `/v1/realtime/transcription_sessions` `/v1/realtime/translations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `Code Interpreter tool` `File Search` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` |

179| Australia\* | `au.api.openai.com` | Yes | No | Yes | Text, Audio, Voice, Image | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/files` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `/v1/vector_stores` `Code Interpreter tool` `File Search` `File Uploads` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` | None |179| Australia\* | `au.api.openai.com` | Yes | No | Yes | Text, Audio, Voice, Image | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/files` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `/v1/vector_stores` `Code Interpreter tool` `File Search` `File Uploads` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` | None |

180| Canada\* | `ca.api.openai.com` | Yes | No | Yes | Text, Audio, Voice, Image | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/files` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `/v1/vector_stores` `Code Interpreter tool` `File Search` `File Uploads` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` | None |180| Canada\* | `ca.api.openai.com` | Yes | No | Yes | Text, Audio, Voice, Image | `/v1/audio/transcriptions, /v1/audio/translations, /v1/audio/speech` `/v1/batches` `/v1/chat/completions` `/v1/embeddings` `/v1/files` `/v1/fine_tuning/jobs` `/v1/images/edits` `/v1/images/generations` `/v1/moderations` `/v1/responses` `/v1/responses File Search` `/v1/responses Web Search` `/v1/vector_stores` `Code Interpreter tool` `File Search` `File Uploads` `Remote MCP server tool` `Scale Tier` `Structured Outputs (excluding schema)` `Supported input modalities` | None |

192#### API Endpoint, tool and model support192#### API Endpoint, tool and model support

193 193

195| -------------------------------------------------------------------- | ---------------- | ----------------------------------------- | --------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------- |195| -------------------------------------------------------------------- | ---------------- | ----------------------------------------- | --------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------- |

197| `/v1/batches` | Batches | All listed regions | United States, Europe (EEA + Switzerland) | `gpt-5.5-pro-2026-04-23`, `gpt-5.4-pro-2026-03-05`, `gpt-5.2-pro-2025-12-11`, `gpt-5-pro-2025-10-06`, `gpt-5.6-sol`, `gpt-5.6-terra`, `gpt-5.6-luna`, `gpt-5.5-2026-04-23`, `gpt-5.4-2026-03-05`, `gpt-5-2025-08-07`, `gpt-5.4-mini-2026-03-17`, `gpt-5.4-nano-2026-03-17`, `gpt-5.2-2025-12-11`, `gpt-5.1-2025-11-13`, `gpt-5-mini-2025-08-07`, `gpt-5-nano-2025-08-07`, `gpt-4.1-2025-04-14`, `gpt-4.1-mini-2025-04-14`, `gpt-4.1-nano-2025-04-14`, `o3-2025-04-16`, `o4-mini-2025-04-16`, `o1-pro`, `o1-pro-2025-03-19`, `o3-mini-2025-01-31`, `o1-2024-12-17`, `o1-mini-2024-09-12`, `o1-preview`, `gpt-4o-2024-11-20`, `gpt-4o-2024-08-06`, `gpt-4o-mini-2024-07-18`, `gpt-4-turbo-2024-04-09`, `gpt-4-0613`, `gpt-3.5-turbo-0125` | None | — |197| `/v1/batches` | Batches | All listed regions | United States, Europe (EEA + Switzerland) | `gpt-5.5-pro-2026-04-23`, `gpt-5.4-pro-2026-03-05`, `gpt-5.2-pro-2025-12-11`, `gpt-5-pro-2025-10-06`, `gpt-5.6-sol`, `gpt-5.6-terra`, `gpt-5.6-luna`, `gpt-5.5-2026-04-23`, `gpt-5.4-2026-03-05`, `gpt-5-2025-08-07`, `gpt-5.4-mini-2026-03-17`, `gpt-5.4-nano-2026-03-17`, `gpt-5.2-2025-12-11`, `gpt-5.1-2025-11-13`, `gpt-5-mini-2025-08-07`, `gpt-5-nano-2025-08-07`, `gpt-4.1-2025-04-14`, `gpt-4.1-mini-2025-04-14`, `gpt-4.1-nano-2025-04-14`, `o3-2025-04-16`, `o4-mini-2025-04-16`, `o1-pro`, `o1-pro-2025-03-19`, `o3-mini-2025-01-31`, `o1-2024-12-17`, `gpt-4o-2024-11-20`, `gpt-4o-2024-08-06`, `gpt-4o-mini-2024-07-18`, `gpt-4-turbo-2024-04-09`, `gpt-4-0613`, `gpt-3.5-turbo-0125` | None | — |

198| `/v1/chat/completions` | Chat Completions | All listed regions | United States, Europe (EEA + Switzerland), United Arab Emirates | `gpt-5.6-sol`, `gpt-5.6-terra`, `gpt-5.6-luna`, `gpt-5.5-2026-04-23`, `gpt-5.4-2026-03-05`, `gpt-5.4-mini-2026-03-17`, `gpt-5.4-nano-2026-03-17`, `gpt-5.2-2025-12-11`, `gpt-5.1-2025-11-13`, `gpt-5-2025-08-07`, `gpt-5-mini-2025-08-07`, `gpt-5-nano-2025-08-07`, `gpt-5-chat-latest-2025-08-07`, `gpt-4.1-2025-04-14`, `gpt-4.1-mini-2025-04-14`, `gpt-4.1-nano-2025-04-14`, `o3-mini-2025-01-31`, `o3-2025-04-16`, `o4-mini-2025-04-16`, `o1-2024-12-17`, `o1-mini-2024-09-12`, `o1-preview`, `gpt-4o-2024-11-20`, `gpt-4o-2024-08-06`, `gpt-4o-mini-2024-07-18`, `gpt-4-turbo-2024-04-09`, `gpt-4-0613`, `gpt-3.5-turbo-0125` | United Arab Emirates: `gpt-5.2-2025-12-11`, `gpt-4.1-2025-04-14` | — |198| `/v1/chat/completions` | Chat Completions | All listed regions | United States, Europe (EEA + Switzerland), United Arab Emirates | `gpt-5.6-sol`, `gpt-5.6-terra`, `gpt-5.6-luna`, `gpt-5.5-2026-04-23`, `gpt-5.4-2026-03-05`, `gpt-5.4-mini-2026-03-17`, `gpt-5.4-nano-2026-03-17`, `gpt-5.2-2025-12-11`, `gpt-5.1-2025-11-13`, `gpt-5-2025-08-07`, `gpt-5-mini-2025-08-07`, `gpt-5-nano-2025-08-07`, `gpt-4.1-2025-04-14`, `gpt-4.1-mini-2025-04-14`, `gpt-4.1-nano-2025-04-14`, `o3-mini-2025-01-31`, `o3-2025-04-16`, `o4-mini-2025-04-16`, `o1-2024-12-17`, `gpt-4o-2024-11-20`, `gpt-4o-2024-08-06`, `gpt-4o-mini-2024-07-18`, `gpt-4-turbo-2024-04-09`, `gpt-4-0613`, `gpt-3.5-turbo-0125` | United Arab Emirates: `gpt-5.2-2025-12-11`, `gpt-4.1-2025-04-14` | — |

209| `/v1/realtime (legacy previews)` | Realtime | United States | United States | `gpt-4o-realtime-preview-2024-12-17`, `gpt-4o-realtime-preview-2024-10-01`, `gpt-4o-mini-realtime-preview-2024-12-17` | None | — |209| `/v1/responses` | Responses | All listed regions | United States, Europe (EEA + Switzerland), United Arab Emirates | `gpt-5.5-pro-2026-04-23`, `gpt-5.4-pro-2026-03-05`, `gpt-5.2-pro-2025-12-11`, `gpt-5-pro-2025-10-06`, `gpt-5.6-sol`, `gpt-5.6-terra`, `gpt-5.6-luna`, `gpt-5.5-2026-04-23`, `gpt-5.4-2026-03-05`, `gpt-5-2025-08-07`, `gpt-5.4-mini-2026-03-17`, `gpt-5.4-nano-2026-03-17`, `gpt-5.2-2025-12-11`, `gpt-5.1-2025-11-13`, `gpt-5-mini-2025-08-07`, `gpt-5-nano-2025-08-07`, `gpt-4.1-2025-04-14`, `gpt-4.1-mini-2025-04-14`, `gpt-4.1-nano-2025-04-14`, `o3-2025-04-16`, `o4-mini-2025-04-16`, `o1-pro`, `o1-pro-2025-03-19`, `o3-mini-2025-01-31`, `o1-2024-12-17`, `gpt-4o-2024-11-20`, `gpt-4o-2024-08-06`, `gpt-4o-mini-2024-07-18`, `gpt-4-turbo-2024-04-09`, `gpt-4-0613`, `gpt-3.5-turbo-0125` | United Arab Emirates: `gpt-5.2-2025-12-11`, `gpt-4.1-2025-04-14` | — |

210| `/v1/responses` | Responses | All listed regions | United States, Europe (EEA + Switzerland), United Arab Emirates | `gpt-5.5-pro-2026-04-23`, `gpt-5.4-pro-2026-03-05`, `gpt-5.2-pro-2025-12-11`, `gpt-5-pro-2025-10-06`, `gpt-5.6-sol`, `gpt-5.6-terra`, `gpt-5.6-luna`, `gpt-5.5-2026-04-23`, `gpt-5.4-2026-03-05`, `gpt-5-2025-08-07`, `gpt-5.4-mini-2026-03-17`, `gpt-5.4-nano-2026-03-17`, `gpt-5.2-2025-12-11`, `gpt-5.1-2025-11-13`, `gpt-5-mini-2025-08-07`, `gpt-5-nano-2025-08-07`, `gpt-5-chat-latest-2025-08-07`, `gpt-4.1-2025-04-14`, `gpt-4.1-mini-2025-04-14`, `gpt-4.1-nano-2025-04-14`, `o3-2025-04-16`, `o4-mini-2025-04-16`, `o1-pro`, `o1-pro-2025-03-19`, `computer-use-preview`, `o3-mini-2025-01-31`, `o1-2024-12-17`, `o1-mini-2024-09-12`, `o1-preview`, `gpt-4o-2024-11-20`, `gpt-4o-2024-08-06`, `gpt-4o-mini-2024-07-18`, `gpt-4-turbo-2024-04-09`, `gpt-4-0613`, `gpt-3.5-turbo-0125` | United Arab Emirates: `gpt-5.2-2025-12-11`, `gpt-4.1-2025-04-14` | computer-use-preview is supported only in the US and Europe. |

230 229

231#### /v1/responses230#### /v1/responses

232 231

233- computer-use-preview snapshots are only supported for US/EU.

234- Cannot set background=True in EU region.232- Cannot set background=True in EU region.

235- [Extended prompt caching](https://developers.openai.com/api/docs/guides/prompt-caching#prompt-cache-retention) in regions that do not support Regional processing may require that OpenAI process and temporarily store Customer Content outside of the Region to deliver the services.233- [Extended prompt caching](https://developers.openai.com/api/docs/guides/prompt-caching#prompt-cache-retention) in regions that do not support Regional processing may require that OpenAI process and temporarily store Customer Content outside of the Region to deliver the services.

236 234

Documentation 2026-07-30 23:58 UTC to 2026-07-31 21:03 UTC