Last reviewed
May 9, 2026
Sources
No citations yet
Review status
Needs citations
Revision
v2 ยท 2,500 words
Add missing citations, update stale details, or suggest a clearer explanation.
See also: ChatGPT Plugins, ChatGPT Plugin Categories and Photography
Photography ChatGPT Plugins were a small group of third-party extensions inside ChatGPT that helped the chatbot work with photographs, photographic filters, and prompts used to produce photography-style imagery in external image models. The category sat inside the broader ChatGPT Plugins program operated by OpenAI, which opened in beta on March 23, 2023, expanded to all ChatGPT Plus subscribers on May 12, 2023, and was wound down between March 19, 2024 and April 9, 2024.[1][2][8]
The Photography tab was one of the thinner categories in the plugin store. It overlapped heavily with the Art ChatGPT Plugins and Design ChatGPT Plugins tabs, and the boundary was often porous. The most directly photographic entry was Polarr, a filter search plugin from a venture-backed photo editor of the same name. Around it sat a small ring of resolution upscalers, image search wrappers, OCR tools, photographic prompt formatters, and NASA image browsers that were promoted as photography helpers in third-party roundups.[14][15][16]
OpenAI announced ChatGPT plugins on March 23, 2023, as tools that help ChatGPT access up to date information, run computations, or use third party services. The launch wave included twelve partner integrations and two first-party plugins for browsing and a code interpreter. None of the original twelve were focused on photography.[1]
Access opened in alpha to ChatGPT Plus subscribers and waitlisted developers. On May 12, 2023, OpenAI announced that web browsing and plugins would become available to all Plus subscribers the following week, and the plugin store opened broadly on May 19, 2023. By mid-May, third-party roundups counted roughly seventy approved plugins; only a handful were filed as photography.[2][14] Polarr appeared in the canonical list of approved plugins published on May 14, 2023.[14]
During summer 2023, several photo plugins reached the store. MixerBox shipped a family of plugins including PhotoMagic, an upscaler, and ImageGen, a generation wrapper.[6] Photorealistic, a Midjourney prompt formatter focused on photographic detail, was widely covered by reviewers.[5][16] Pixellow, an image captioning plugin, and SceneXplain by Jina AI, a multilingual image description tool, both began rolling out to Plus accounts in mid-2023.[10][11]
On October 19, 2023, DALL-E 3 was released in beta inside ChatGPT Plus and Enterprise as a built-in feature. On November 6, 2023, at the first OpenAI DevDay, Custom GPTs were unveiled. The GPT Store launched on January 10, 2024. New plugin installs and conversations were disabled on March 19, 2024, and existing plugin conversations stopped working on April 9, 2024.[3][4][7][8]
| Date | Event |
|---|---|
| March 23, 2023 | ChatGPT plugins announced; no photography launch partners |
| May 12, 2023 | Plugins rolled out broadly to ChatGPT Plus |
| May 14, 2023 | Polarr present on the canonical approved-plugin list |
| Mid 2023 | MixerBox PhotoMagic, Photorealistic, Pixellow, SceneXplain reach Plus accounts |
| October 19, 2023 | DALL-E 3 integrated natively into ChatGPT |
| November 6, 2023 | GPTs unveiled at OpenAI DevDay |
| January 10, 2024 | GPT Store opens |
| March 19, 2024 | New plugin installs and conversations disabled |
| April 9, 2024 | Existing plugin conversations stopped working |
All photography plugins shared the standard plugin architecture. Each plugin advertised itself through a manifest file and an OpenAPI specification that ChatGPT consumed at install time. GPT-4 decided whether to call one or more installed plugins, sent structured arguments to the plugin endpoint, received a response, and incorporated it into the reply. Users could enable up to three plugins per conversation, and plugin features were limited to ChatGPT Plus subscribers running on GPT-4.[1][17]
Photography plugins fell into a few recurring patterns:
Many plugins blurred categories. PhotoMagic was promoted as photography but is also reasonable to file under Design or Productivity. Photorealistic and Imgenic could equally have lived in Art. ChatOCR was filed as Productivity in some directories and Photography in others. The boundaries reflected developer self-classification more than any strict OpenAI taxonomy.[15]
The plugins below were repeatedly cited as photography-category entries in third-party directories during the plugin beta. All claims describe behaviour during that beta period; none of these plugins remain installable through ChatGPT today. This wiki only lists plugins that can be cross-referenced to multiple credible sources. Plugins that appeared in a single roundup without independent confirmation are omitted.
Polarr was the most directly photographic plugin in the plugin store. It let ChatGPT search Polarr's catalogue of user generated filters, returning a text description of a recommended filter and a preview link. Typical prompts described a desired effect, such as a vintage colour cast, a moody black and white look, a sunset enhancement, or a face retouch, and the plugin replied with a matching filter that the user could open inside the Polarr photo editor.[15][18][19]
Polarr is a venture-backed photo and video editor founded in August 2014 by Borui Wang and Derek Yan and headquartered in San Jose, California. The company raised an 11.5 million United States dollar Series A round led by Threshold Ventures in March 2019.[20] The ChatGPT plugin connected to that catalogue rather than running any editing locally; the manifest in community-maintained plugin store specs identified it as version 0.4.8 with a single search endpoint at api.polarr.co/ml/adjustment/assets/gpt-plugin-search.[19]
MixerBox PhotoMagic was an enhancement plugin that accepted an image URL and returned an upscaled version with cleaner detail. It was promoted as a way to repair compression artefacts, sharpen scanned documents, restore detail in compressed photographs, and improve the apparent resolution of forwarded camera-roll images. Users could specify a scaling factor. Hair, skin, and fine texture were called out in the marketing as targets where the upscaling model was tuned to retain detail rather than smooth aggressively.[6] MixerBox shipped a large family of ChatGPT plugins from one developer account, which made PhotoMagic unusually visible in directory roundups despite using a generic upscaler.[15]
Photorealistic was a prompt-crafting plugin that turned short user descriptions into long form, photography-style prompts intended for Midjourney. The keyword photorealistic in a chat triggered the plugin, which returned two engineered prompts. The output usually specified a camera body, often a Canon EOS 5D Mark IV or a Canon EOS R5, a lens, a lighting condition, a colour palette, and a cinematic style. Users copied the output into Midjourney or Stable Diffusion running outside ChatGPT.[5][16] The plugin sat at the intersection of photography and prompt engineering; it produced no images itself.
Imgenic was a second photographic prompt plugin focused on Midjourney. It exposed two operations: searchPrompts for browsing example prompts, and rankPrompts for evaluating prompt quality. Promotional copy emphasised eight kilo resolution targets, equipment such as Canon EOS R5 bodies, golden hour lighting, and named photography styles such as fine art and commercial. Imgenic did not call Midjourney; the user pasted the resulting prompt themselves. The plugin was documented in directories from September 2023.[12]
Pixellow was an image captioning and description plugin built by Ceylon AI. It analysed an attached image and returned structured text covering objects, dominant colours, visible writing, location cues, mood, recognised people and animals, the main subject, and a draft caption. The endpoint domain was pixellow.ceylon.ai.[10]
SceneXplain was an image storytelling plugin from Jina AI. It accepted an image URL or a base64 encoded image and returned a long-form description that captured multiple objects, interactions, and contextual elements. SceneXplain was promoted as going beyond simple captioning produced by older models such as CLIP, BLIP, or BLIP2, with a focus on multilingual output and richer scene reasoning. The plugin built on Jina AI's standalone product at scenex.jina.ai.[11]
ChatOCR accepted scans, photographs of documents, and handwritten notes, and returned the extracted text inside the chat. Directories filed ChatOCR variously under Productivity and Photography because it processed photographs of documents but was usually used for archival and transcription work. It appeared on multiple May 2023 roundup lists.[14][15]
Astrodaily was a NASA image search plugin built by an independent developer with the GitHub handle 360macky. It returned the Astronomy Picture of the Day and could search the NASA Image Library by keyword. Output was rendered inline as Markdown images. The plugin's API was hosted at api.360macky.com.[13]
Space Photo Explorer was a second NASA-backed photography plugin. It exposed three datasets: the Astronomy Picture of the Day, a NASA Image Library search, and a Mars Rover photo retrieval endpoint. For Mars photographs the user could specify a sol number, a camera identifier, an Earth date, and a rover name. The manifest was hosted at photoexplorer.space.[9]
Sponsored.pro was a product photography prompt generator aimed at marketers. It produced detailed prompts for product and lifestyle photography that the user fed into a separate image model. Prompts could be customised with colours, props, and named photographic styles.[15]
Directories sometimes mentioned other photography-tagged plugins, including Image Editor for resize, crop, and blur; ImageSearch for fetching photographs from Unsplash; WebRewind for Internet Archive snapshots; and Earth for generated map images. These returned image URLs but their primary purpose was tooling around hosted images rather than photographic work.[15]
The Photography tab was unusually porous, like the Art ChatGPT Plugins tab next to it. The table below summarises common boundary cases that directories disagreed about.
| Plugin | Filed as photography because | More precise category |
|---|---|---|
| Photorealistic | Generates camera and lens prompts | Prompt Engineering |
| Imgenic | Generates camera and lens prompts | Prompt Engineering |
| ChatOCR | Reads text from photographs | Productivity |
| Sponsored.pro | Builds product photography prompts | Marketing |
| Image Editor | Modifies hosted images | Design |
| ImageSearch | Returns photographs from Unsplash | Search |
| WebRewind | Returns historical website screenshots | Web Archive |
| Earth | Returns generated map imagery | Maps |
| MixerBox PhotoMagic | Upscales photographs | Photography or Productivity |
The most consequential boundary issue was DALL-E. DALL-E 2 had been OpenAI's dominant text to image system when plugins launched, but it was never offered as a third-party plugin. DALL-E 3 was integrated into ChatGPT directly on October 19, 2023, accessed through a model picker rather than the plugin store. After October 2023, anyone who wanted to create photographic-looking imagery inside ChatGPT used DALL-E 3 directly, leaving the third-party photography plugins to occupy a thin slice of demand: filter recommendations, upscaling, prompt preparation for non-OpenAI image models, and OCR or analysis of an existing photograph.[3]
Several structural factors kept the photography tab smaller than peers such as Productivity or Travel:
The direct successor to in-chat photographic imagery was native DALL-E 3 integration. Once DALL-E 3 was available inside Plus and Enterprise from October 2023, users could ask ChatGPT for a photograph-style image without installing anything. Subsequent updates added vision capabilities so the model could analyse photographs without a separate captioning plugin, replacing much of what Pixellow and SceneXplain had offered.[3]
Custom GPTs are user-created versions of ChatGPT that combine instructions, optional knowledge files, and Actions, the direct successor to plugins. The GPT Store launched on January 10, 2024 with photography-themed GPTs visible from day one. Many photography plugin developers ported their offerings to GPTs that wrapped an existing API behind a system prompt and a small knowledge file; the Polarr listing in third-party GPT directories continued under the GPTs framework after the plugin beta closed.[7][18] Users who wanted controls beyond what ChatGPT offered returned to standalone tools such as Adobe Lightroom, Topaz Photo AI, and Midjourney workflows on Discord.
The photography plugin category is now framed as a transitional product, like its sibling Art ChatGPT Plugins and Design ChatGPT Plugins categories. It was small at launch, never grew large, and was hollowed out when DALL-E 3 and ChatGPT vision features absorbed most casual use and Custom GPTs replaced plugins as the default authoring surface. Even so, the category proved that a large language model could orchestrate calls to a filter catalogue, an upscaler, a captioner, and a prompt formatter from inside a single chat.