Photography ChatGPT Plugins

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. They are deprecated and no longer usable: OpenAI closed the plugin store to new installs and new conversations on March 19, 2024, and existing plugin conversations stopped working on April 9, 2024.^[1][2][8] The category had sat inside the broader ChatGPT Plugins program, which opened in beta on March 23, 2023 and expanded to all ChatGPT Plus subscribers on May 12, 2023.^[1][2] OpenAI replaced plugins with a chain of successor surfaces: Custom GPTs and Actions in the GPT Store (from November 2023), ChatGPT connectors (2025), the Model Context Protocol (MCP, which OpenAI adopted on March 26, 2025), and Apps in ChatGPT built with the Apps SDK (announced October 6, 2025).^{[4][7][22][24]}

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]

When were photography ChatGPT plugins available?

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. Anthropic then introduced MCP on November 25, 2024, OpenAI adopted it on March 26, 2025, and Apps in ChatGPT and the Apps SDK were announced on October 6, 2025.^{[3][4][7][8][21][22][24]}

Scroll sideways for more →

What did photography ChatGPT plugins do?

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:

• filter search, querying a remote catalogue of photographic filters. Polarr was the canonical example.
• image enhancement, accepting an image URL and returning an upscaled version. MixerBox PhotoMagic was the most reviewed entry.
• photographic prompt drafting, expanding a short description into a long form prompt for Midjourney or Stable Diffusion. Photorealistic and Imgenic were the leading examples.
• image description and captioning, returning analytical text covering subject, mood, dominant colour, and context. Pixellow and SceneXplain were widely cited.
• image catalogue search, retrieving photographs from public archives such as the NASA Image Library. Astrodaily and Space Photo Explorer wrapped NASA APIs.
• document and photo OCR, extracting text from scanned PDFs, photographs, and handwriting. ChatOCR was the most cited entry.
• product photography prompt building, proposing product photography prompts for marketing assets. Sponsored.pro was the typical example.

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]

What were the notable photography ChatGPT plugins?

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

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

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

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

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

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

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

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

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

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

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]

plugins commonly listed but not separately documented

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]

How did photography plugins relate to art, design, and built-in features?

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.

Scroll sideways for more →

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]

Why did the photography plugin category stay small?

Several structural factors kept the photography tab smaller than peers such as Productivity or Travel:

• first party competition. Once DALL-E was inside ChatGPT directly, third-party image generation plugins lost most of their reason to exist.
• plugin pricing. Image inference and upscaling were costly to operate. Several plugins relied on per-user quotas tied to a developer-side account.
• install friction. Users could enable only three plugins per conversation. Combining a filter search, an upscaler, and a prompt formatter exhausted that budget.
• discoverability. The plugin store sat behind a model picker. The Photography tab was rarely surfaced in OpenAI marketing.
• a richer standalone market. Photography software had a deep set of standalone tools, including Adobe Lightroom and the Polarr editor itself. A plugin that lived inside a chat and could not modify a local file was unlikely to displace those workflows.

What replaced photography ChatGPT plugins?

Plugins were retired in favour of a chain of successor surfaces, each broader than the last: native image generation, then Custom GPTs and Actions, then connectors, then the Model Context Protocol, and finally Apps in ChatGPT.

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. Actions call an external API from an OpenAPI schema in the same way a plugin manifest did, so Custom GPTs inherited the plugin integration model rather than discarding it. 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.

Through 2025 OpenAI added ChatGPT connectors, which let ChatGPT read from a user's own external sources such as Google Drive, SharePoint, Dropbox, and Box instead of installing third-party apps. Connectors reached Pro users in June 2025 and were later folded into the broader apps framework.^[23]

The connector and app model was standardised on the Model Context Protocol (MCP), an open standard that Anthropic introduced on November 25, 2024 for connecting AI systems to external tools and data.^[21] OpenAI adopted MCP on March 26, 2025. CEO Sam Altman wrote that "People love MCP and we are excited to add support across our products," with support arriving first in the Agents SDK and then the ChatGPT desktop app and Responses API.^[22]

The current authoring surface is Apps in ChatGPT, announced with the Apps SDK at OpenAI DevDay on October 6, 2025. The Apps SDK builds on MCP and lets developers define both the logic and the interface of an app that runs inside a conversation; launch partners included Spotify, Canva, Zillow, Figma, Coursera, Expedia, and Booking.com.^[24] For photography specifically this closes the loop the plugin store opened: a filter catalogue, an upscaler, or a prompt tool that once shipped as a Photography plugin would today ship as an MCP-based app or a Custom GPT, while most casual image creation and analysis is handled natively by DALL-E 3 and ChatGPT vision.

Why do photography ChatGPT plugins matter?

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, a role now carried by MCP-based Apps in ChatGPT. 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, the same pattern that later successor surfaces generalised.

See also

• ChatGPT Plugins
• ChatGPT Plugin Categories
• Photography
• Art ChatGPT Plugins
• Design ChatGPT Plugins
• Custom GPTs
• GPT Store
• Model Context Protocol
• OpenAI Apps SDK
• ChatGPT
• OpenAI
• GPT-4
• Prompt Engineering

References

1. OpenAI. *ChatGPT plugins*. March 23, 2023. https://openai.com/index/chatgpt-plugins/ ↩
2. OpenAI. *ChatGPT Release Notes*. https://help.openai.com/en/articles/6825453-chatgpt-release-notes ↩
3. OpenAI. *DALL-E 3 is now available in ChatGPT Plus and Enterprise*. October 19, 2023. https://openai.com/index/dall-e-3-is-now-available-in-chatgpt-plus-and-enterprise/ ↩
4. OpenAI. *Introducing GPTs*. November 6, 2023. https://openai.com/index/introducing-gpts/ ↩
5. Sercan Tarakcioglu. *Photorealistic: A Chat-GPT plugin*. Medium. https://medium.com/@_seca_/photorealistic-a-chat-gpt-plugin-to-generate-photorealistic-images-on-midjourney-7b07228ec42c ↩
6. MixerBox. *MixerBox PhotoMagic*. https://www.mixerbox.com/blog/photomagic ↩
7. VentureBeat. *OpenAI launches GPT Store*. January 10, 2024. https://venturebeat.com/ai/openai-launches-gpt-store-but-revenue-sharing-is-still-to-come ↩
8. CloudWatt. *ChatGPT is Removing Plugins From March 19, 2024*. https://cloudwatt.com/chatgpt-is-removing-plugins-from-march-19-2024/ ↩
9. Space Photo Explorer plugin manifest. https://github.com/sisbell/chatgpt-plugin-store/blob/main/manifests/photoexplorer.space.json ↩
10. Pixellow ChatGPT Plugin listing. https://gptstore.ai/plugins/pixellow-ceylon-ai ↩
11. Jina AI. *SceneXplain*. https://jina-ai-gmbh.ghost.io/podcast/scenexplain-unleash-advanced-image-captioning-storytelling/ ↩
12. Imgenic listing. https://plugin.surf/plugin/imgenic ↩
13. Marcelo Arias. *Create a ChatGPT Plugin to retrieve NASA images*. https://dev.to/marcelo_earth/create-a-chatgpt-plugin-to-retrieve-nasa-images-45pk ↩
14. Orren Prunckun. *Approved ChatGPT Plugins As Of 14 May 2023*. https://orren.medium.com/a-list-of-all-approved-chatgpt-plugins-as-of-14-may-2023-4c60afbfdcfc ↩
15. ScriptByAI. *30+ ChatGPT Photo Plugins*. https://www.scriptbyai.com/chatgpt-photo-plugins/ ↩
16. ROIhacks. *Photorealistic Plugin Review*. https://roihacks.com/photorealistic-chatgpt-plugin/ ↩
17. CustomGPT. *How To Install ChatGPT Plugins*. https://customgpt.ai/how-to-install-chatgpt-plugins-customgpt/ ↩
18. Polarr listing. https://gptstore.ai/plugins/polarr-co ↩
19. Polarr OpenAI Plugin spec. https://github.com/sisbell/chatgpt-plugin-store/blob/main/specs/polarr.co.yaml ↩
20. Polarr. Wikipedia. https://en.wikipedia.org/wiki/Polarr ↩
21. Anthropic. *Introducing the Model Context Protocol*. November 25, 2024. https://www.anthropic.com/news/model-context-protocol ↩
22. TechCrunch. *OpenAI adopts rival Anthropic's standard for connecting AI models to data*. March 26, 2025. https://techcrunch.com/2025/03/26/openai-adopts-rival-anthropics-standard-for-connecting-ai-models-to-data/ ↩
23. TechCrunch. *ChatGPT introduces meeting recording and connectors for Google Drive, Box, and more*. June 4, 2025. https://techcrunch.com/2025/06/04/chatgpt-introduces-meeting-recording-and-connectors-for-google-drive-box-and-more/ ↩
24. OpenAI. *Introducing apps in ChatGPT and the new Apps SDK*. October 6, 2025. https://openai.com/index/introducing-apps-in-chatgpt/ ↩