What we built in FOTOhub's first year

Summary

On September 11, 2025, FOTOhub went live as an AI image generator built by a small team in Poland. Twelve months later it is a Creative AI hub that serves more than 1 million users in 46 countries, runs 226 AI models for image, video, audio, chat, text, 3D and virtual try-on behind one account and one balance, and ships a developer platform with 314 REST endpoints, official SDKs, an MCP server and hourly GPU capacity. Along the way the company raised $8.6 million, joined the Amazon Web Services partner portfolio, trained and launched its first proprietary model, and was valued at $180 million. This is the full account of the first year: what was built, in what order, what it runs on, what it costs and where it goes next.

September 11, 2025: one image generator

The first version of FOTOhub did one thing. You typed a prompt, picked a model, and got an image. There was a gallery, a subscription and a very short roadmap. It was a good image generator, and that was the point: prove that people would pay a Polish startup for access to frontier models, then earn the right to build the rest.

They did pay. Within weeks the product had a registration flow, paid plans and a moderation pipeline, and the questions users asked started to change shape. Can I animate this? Can I add a voice? Can I get this into my shop? Can I call this from code? Every one of those questions became a product in the year that followed.

The company was incorporated in Delaware with engineering in Poland, and it kept that arrangement: an American company with a Polish engineering culture, shipping at a pace that surprised people on both sides of the Atlantic.

The numbers, twelve months later

Everything below is counted from the platform as it runs today, September 2026.

	September 2025	September 2026
Users	first paying customers	more than 1,000,000 in 46 countries
AI models	one generation family	226 across eight categories
Media	image	image, video, audio, chat, text, 3D, try-on
Studios	one generator	photo editor, video studio, Chat Live, Canvas, Flows, FH Shorts, UGC Ads, Story Studio, 3D Studio, Orbit Designer, Presentation Studio, try-on, photo restoration
Developer platform	none	314 REST endpoints in 38 modules, Python and TypeScript SDKs, CLI, n8n node, 28 MCP tools, webhooks, S3-compatible storage, hourly CPU and GPU
Serverless functions	a handful	301
Backend services	one	33
Documentation	none	115 pages at docs.fotohub.app
Proprietary models	none	IDA Q 1.0 for image, IDA Voice for speech
Capital raised	bootstrapped	$8.6 million
Valuation	none	$180 million, Q2 2026
Cloud	rented capacity	Amazon Web Services partner portfolio, own GPU fleet, 16 regions

One of those rows deserves a note. The jump from a handful of serverless functions to 301 is the visible trace of every feature that had to be wired to billing, storage, moderation and notifications before it could be called finished.

!Creators filming for FH Shorts and UGC Ads

What FOTOhub is now: one account, every medium

The simplest description of FOTOhub in 2026 is this: one account, one balance of credits, and everything a creative project needs to move from idea to published asset without leaving the platform.

An image made in the generator can be animated in the video studio, voiced in the audio studio, cut into vertical clips in FH Shorts, placed on a design in Canvas, packaged as a product ad in UGC Ads and published to a shop through a plugin, and at no step does the user download a file, convert a format or sign in somewhere else. Every asset lands in one library with unlimited storage on paid plans, and every operation shows its price before it runs. Every model, whether it came from a global provider or from FOTOhub's own training runs, is reached through the same interface.

That is what the phrase Creative AI hub means in practice. Not a bundle of tools with a shared login, but a single system where the output of any model is the input of any other.

226 models, one orchestration layer

The catalogue today, by category:

• 63 image models, among them Seedream 5.0 and 4.5, GPT Image, Imagen and FOTOhub's own IDA Q 1.0
• 65 video models, among them Seedance 2.5, Luma Ray 3.2, Veo and Sora
• 30 audio operations covering text to speech, music, sound effects, stem separation, mastering, transcription and voice cloning
• 35 chat models across the GPT, Claude, Gemini, DeepSeek and Nova families
• 20 text models for copy, scripts and translation
• 3D generation, virtual try-on and a set of utility models for upscaling, background removal and object cleanup

The number that matters more than 226 is one: the orchestration layer that sits behind all of them. Every provider has its own request format, its own error codes, its own rate limits and its own habit of deprecating things on short notice. FOTOhub absorbs all of that in one place. A model that degrades reports itself unavailable instead of accepting work it cannot finish. Streaming starts from the first token on every provider that supports it, including through the public API. A fix to one provider's quirk benefits every studio at once.

Gabriel, the platform's orchestrator, is the user-facing side of that layer. Ask it for a result rather than a model and it picks the model, sets the parameters and routes the job. Developers can call the same routing logic from the API and from MCP.

Each medium, in depth

Image. The generator was rebuilt in August as a single surface for every image model in the catalogue: one prompt box, one reference-image tray, one set of aspect ratios, and the model as a choice rather than a separate product. A job can return up to 15 images. Reference images, masks and inpainting work the same way across models that support them. On Professional and Business, Seedream 5.0, Seedream 4.5 and GPT Image 1.5 run without credit charges inside a daily fair-use ceiling, so the daily grind of iteration stops touching the balance.

Video. Sixty-five models, and the operations around them: extension, keyframes, image to video, subtitles, speed and stabilization. Seedance 2.5 generates up to 30 seconds in one pass from as many as 50 references, so a cast, a location and a soundtrack can be fixed in a single request. Luma Ray 3.2 renders 5 or 10 seconds up to 1080p with HDR and EXR output and ships with edit and reframe variants. Every video job runs asynchronously with a job id that the dashboard, the API and MCP all understand.

Audio. Thirty operations under the IDA Y umbrella: narration with IDA Voice, full music tracks, sound effects, stem separation, mastering, transcription and voice cloning, which requires a consent check before a voice can be cloned. Audio was the quietest launch of the year, and one many creators only find when a clip they have just cut needs a voiceover.

Chat and text. Thirty-five chat models and 20 text models in one workspace, streaming from the first token on every provider that supports it. Reasoning models expose their thinking where the provider permits. Since September the same workspace can call tools on external MCP servers, which turns a chat into a place where work gets filed.

3D and try-on. Image to 3D and text to 3D, PBR texturing for the Pro pipeline, and a mesh toolbox for analysis, remeshing, repair, format conversion and UV unwrapping, with a dedicated path for 3D printing. Virtual try-on renders apparel on a model in around ten seconds and feeds the e-commerce plugins directly.

Our own models: the IDA family

A platform that only resells other companies' models is a reseller. In July 2026 FOTOhub crossed that line with IDA Q 1.0, its first proprietary image generation model, trained and served on the company's own GPU fleet. IDA Q is not the most expensive model in the catalogue; it is the one that anchors the cost of the most common operations, and that anchor is what lets the platform offer selected models without credit charges on the Professional and Business plans.

The audio side of the family is IDA Voice, the speech engine behind narration across the studios and the text_to_speech tool exposed to agents. Music, sound effects and soundtrack generation sit alongside it under the IDA Y umbrella.

Owning models changes the relationship with the rest of the catalogue. When a strong new model ships anywhere in the world, FOTOhub can add it within days without renegotiating its cost structure, because the baseline load already runs on its own hardware.

!From print to screen: the photo editor is where FOTOhub started

The studios: where models become work

Models are components. The studios are what a creator actually opens.

The photo editor. A full editor with layers, masks, inpainting, object and background removal, restoration of old or damaged photos, and upscaling to 4K. It is where the first FOTOhub users lived, and it remains the heart of the platform.

Video studio. Text and image to video across 65 models, extension of existing clips, keyframe control on models that support it, HDR and EXR output on Luma Ray 3.2, plus the operations around a clip: subtitles, speed, stabilization, merging.

Chat Live. A conversational workspace with 35 chat models where the model can build things: generate an image mid-conversation, draft a script, translate it, and since September call external tools through MCP servers the user connects.

Canvas. The board where generated assets become a design: layouts, brand assets, exports ready for a client.

Flows. The same models, running without a person in the loop. A flow chains models into an automation that runs from start to finish without supervision.

FH Shorts. Launched in July with the funding announcement, FH Shorts turns long recordings into vertical clips with hook detection, word-level caption editing and one-click recuts, and it has shipped a major update every month since.

UGC Ads. Launched in August. Product videos with AI presenters, storyboarded from a product photo and a brief, with the cost shown before each render.

Story Studio, 3D Studio, Orbit Designer, Presentation Studio, virtual try-on, photo restoration. Each is a full product with its own workflow, and most of them were added in the last six months. 3D Studio alone covers image to 3D, text to 3D, PBR texturing and mesh repair for print.

Teams. Shared credit pools with roles and per-member limits enforced on the server, for studios of 3 to 50 people.

The developer side: console, API, SDKs, MCP

FOTOhub's second customer is a developer, and the second half of the year was largely built for them.

The developer console at fotohub.app/console is where a project gets its keys, its usage graphs, its webhooks and its storage. The public API at apis.fotohub.app answers 314 endpoints across 38 modules: generation for every medium, editing operations, billing, model discovery, jobs and webhooks with signed payloads. Official SDKs ship on PyPI and npm under the name `fotohub`, a command-line tool covers the same surface for scripts and CI, an n8n node brings the catalogue into low-code automations, and plugins exist for six e-commerce platforms including Shopify, WooCommerce and PrestaShop.

In August the console grew into an infrastructure layer: S3-compatible storage with output routing per model and per region, and hourly CPU and GPU capacity for customers whose needs go past calling an endpoint. In September the MCP server made the whole catalogue callable from inside Claude Desktop, Cursor, Windsurf and VS Code through 28 tools, and Chat Live gained the ability to connect to external MCP servers in return.

Documentation lives at docs.fotohub.app: 115 pages, including a reference for every endpoint and every MCP tool. The default model in every example is Seedream 5.0, because it is the one most developers reach for first.

The developer platform in numbers

REST endpoints	314 across 38 modules at apis.fotohub.app
SDKs	Python and TypeScript, both published as `fotohub`
Tooling	command-line tool, n8n node, plugins for six e-commerce platforms
MCP	28 tools, listed in the official registry as app.fotohub/fotohub
Documentation	115 pages at docs.fotohub.app
Webhooks	signed payloads for the generation lifecycle
Storage	S3-compatible buckets with per-model, per-region output routing
Compute	CPU and GPU capacity by the hour, from the same console
Key controls	scopes, per-key rate limits (60 requests a minute by default, up to 600), one-click revocation
Status	public status page at status.fotohub.app

!Video review on the platform's own rendering pipeline

Infrastructure: from a rented GPU to an AWS partner

The first FOTOhub ran on rented capacity. Today the platform is consolidated on Amazon Web Services, is part of the AWS partner portfolio, and operates its own GPU fleet in Frankfurt and Ireland alongside CPU compute for agents and edge functions for everything that has to be fast.

The consolidation, completed in August, is why regional coverage is now a product feature rather than an infrastructure detail. A customer's generated files can land in a bucket in their own region, under a path layout they define, and a regulator who asks where the data lives gets a clear answer. It is also why the company can sell compute by the hour: the same fleet that renders FOTOhub's own jobs is available to customers who need a GPU for an afternoon.

AWS backed the company early with $100,000 in Activate credits in March, and the relationship deepened in June with the announcement of a joint plan to scale the platform globally.

The year, month by month

Month	What happened
September 2025	FOTOhub launches as an AI image generator
October 2025	Paid plans, registration and moderation pipeline live; first paying customers
January 2026	Developer console, billing engine and API projects built
March 2026	$5.1 million round; Canvas, Flows and Teams launch; TOP 5 on Crunchbase Poland and top 1 percent globally; $100,000 AWS Activate credits; EU AI Act compliance position published
April 2026	600,000 creators; multi-model workflows and infrastructure expansion
June 2026	$2.3 million follow-on from DGE3 Investment; deepened AWS partnership; FOTOhub positioned as a Creative OS
July 2026	Partnership with ByteDance/BytePlus; Seedream 5.0 PRO; API Console, 200+ models and the Gabriel orchestrator; $8.6 million raised to date and FH Shorts launch; IDA Q 1.0, the first proprietary model; Natalia Siwakowska joins as COO
August 2026	AWS consolidation and hourly CPU and GPU; Seedance 2.5 live and Seedream unlimited on Professional and Business; console becomes an infrastructure layer with S3 and webhooks; 226 models under one orchestration layer; UGC Ads launch; FH Shorts word-level editing
September 2026	One plan replaces ten subscriptions; MCP server live in Claude, Cursor and VS Code; first anniversary

Three months in that table, July, August and September, account for more product than most companies ship in a year. That was not an accident. The first half of the year built the machine: billing, orchestration, storage, the console. The second half used it.

Twelve launches that defined the year

Canvas, Flows and Teams (March 13). The first time FOTOhub stopped being a generator with an editor attached. Canvas turned outputs into designs, Flows chained models into automations, and Teams gave studios a shared balance with roles. The platform gained a workspace and a reason for a second person to log in.

Seedream 5.0 PRO (July 10). The model that made the generator a daily tool for professionals. Its arrival set the pattern for every launch since: available inside FOTOhub within days of the provider's release, at the same price list as everything else.

API Console, 200+ models and Gabriel (July 20). The biggest single upgrade of the year. The developer console went public, the catalogue crossed 200 models, and Gabriel began routing jobs to the right model on the user's behalf. It was the moment FOTOhub started describing itself as infrastructure.

FH Shorts (July 22). Announced alongside the $8.6 million funding milestone. Long recordings in, vertical clips out, with hooks detected and captions editable word by word. It has shipped a major update every month since.

IDA Q 1.0 (July 24). The first model FOTOhub trained itself, served on its own GPUs. It changed the company from a platform that resells models to a platform that also makes them, and it anchored the cost structure that makes unlimited generation on selected models possible.

AWS consolidation and hourly compute (August 1). All services moved onto Amazon Web Services, and the fleet that renders FOTOhub's own jobs became available to customers by the hour. Regional storage stopped being a roadmap item.

Seedance 2.5 and unlimited Seedream (August 7). Thirty-second clips from 50 references, plus the decision to stop charging credits for the most used image models on Professional and Business. The pricing article a month later grew out of this change.

Console as an infrastructure layer (August 10). S3-compatible storage with output routing, signed webhooks and the console tabs to manage them. From this point a developer could build a product on FOTOhub without standing up storage or a notification backend of their own.

226 models, one orchestration layer (August 18). The stability release. Fewer visible features than any other launch on this list, and the engineering underneath them all: health monitoring, provider isolation, streaming everywhere, one job model across media.

UGC Ads (August 28). Product videos with AI presenters, storyboarded from a photo and a brief. The first studio built specifically for e-commerce sellers rather than adapted for them.

One plan instead of ten subscriptions (September 2). Not a launch but a statement: the full pricing math against the stack of single-purpose tools, published with real numbers. It set the frame for everything the platform says about price.

MCP server (September 4). FOTOhub inside Claude Desktop, Cursor, Windsurf and VS Code through 28 tools, and external tools inside FOTOhub through Chat Live. The year ended with the platform reachable from places that did not exist as channels when it began.

One price for everything

The pricing philosophy has not changed since day one, only its scale. One subscription, one balance of credits, every model.

Plans start at $8 a month (33 zł) for 300 credits and rise to Business at $175 (700 zł) for 20,000, with a Team plan at $99 (386 zł) per seat and a free tier of 50 credits that needs no card. Professional and Business subscribers generate on Seedream 5.0, Seedream 4.5 and GPT Image 1.5 without spending credits at all, within a daily fair-use ceiling. Buster Packs top up any plan without upgrading it. The same subscription covers the API, the SDKs and the MCP server; there is no separate developer plan.

Last week's article laid out the full math against the stack of single-purpose AI subscriptions a working creator typically pays for. The short version: that stack costs between $97 and $277 a month at list prices. FOTOhub replaces most of it for less than the price of one of them. The complete breakdown is at fotohub.app/news/fotohub-one-plan-instead-of-ten-ai-subscriptions-en.

Who uses FOTOhub

The platform serves six audience segments deliberately, with different entry points for each.

Individual creators and photographers, the original audience, use the editor, the generator and FH Shorts. Agencies and studios run Teams with shared credit pools, brand assets and client-ready exports from Canvas and Presentation Studio. E-commerce sellers generate product photography, try-on imagery and UGC-style video, and push results into their shops through the plugins. Media companies and publishers use Story Studio and the video studio for volume. Developers and startups build on the API, the SDKs and MCP. Educators and students use the free tier and the education pages built in July.

Six segments, one platform. The segments are separated in the interface, in the pricing and in the sales material, but they share the same models, the same storage and the same orchestration layer underneath.

Partners and ecosystem

Amazon Web Services. AWS backed the company with $100,000 in Activate credits in March, welcomed it into the partner portfolio in June with a joint plan to scale the platform globally, and hosts the entire stack today: the GPU fleet, the compute for agents, storage across 16 regions and the console that exposes all of it to customers.

ByteDance and BytePlus. The partnership announced in July brought the Seedance and Seedream families to FOTOhub on launch schedules, including a period in which Seedance 2.0 generated free of charge for every user. Seedream 5.0 is the default model in the documentation and in every code example.

DGE3 Investment. Investor in the March round and the June follow-on, and author of a market report on Creative AI published in July that featured FOTOhub.

Model providers. The catalogue draws on models from the GPT, Claude, Gemini, Seedream, Seedance, Luma, Veo, Sora, DeepSeek and Nova families among others, plus FOTOhub's own IDA line. Provider relationships are what let a new model appear in the catalogue within days of its release.

Two countries, one company. FOTOhub was created in Poland and is incorporated in Wilmington, Delaware. Engineering remains in Poland. The interface is available in English, Polish and ten further languages, and the documentation, the console and the API speak English by default.

Trust, compliance and the boring work

The August release was described internally as the most stable version of the platform to date, and stability is the least glamorous thing to write about. It is also the thing enterprise customers ask about first.

Health monitoring runs continuously under the model catalogue; a degraded provider takes its models out of rotation automatically. Every generation is priced before it runs. Every paid plan carries a commercial license. In March the company published its position on the EU AI Act, ahead of the obligations that apply to general-purpose AI providers and their deployers. Content moderation runs inside the generation pipeline. Teams get server-enforced limits per member, and API keys can be scoped, rate-limited and revoked in one click.

None of that is a feature anyone signs up for. All of it is why the platform could grow from a generator to a hub without breaking under the load.

The company

FOTOhub was founded on September 11, 2025 by Mateusz Ulewicz, Founder and CTO, who designed the orchestration layer, the GPU fleet, the billing system and the IDA model family and remains the architect of the platform. Before FOTOhub he spent two decades in engineering, including work at Amazon Web Services and Nokia. On Crunchbase he ranks in the top 5 globally in his primary category, out of more than 2.1 million profiles.

Natalia Siwakowska joined as Chief Operating Officer in July 2026 to lead operations as the company scaled past its first million users.

The company has raised $8.6 million to date: a seed round, a $5.1 million round in March 2026 and a $2.3 million follow-on from DGE3 Investment in June 2026. Amazon Web Services and DGE3 Investment are among its backers. In Q2 2026 FOTOhub was valued at $180 million. DGE3 published a market report on Creative AI in July featuring FOTOhub. The company is incorporated in Wilmington, Delaware, with engineering in Poland.

Year two

The plan for the second year follows from the first. More IDA models, so that a larger share of the catalogue runs on FOTOhub's own hardware. Agents that can run entire workflows, not single tools, across the studios and the API. MCP coverage for every studio, including Flows, 3D and the editor. Marketplace listings for the e-commerce plugins. More regions, because customers keep asking where their data sits. And a continued monthly cadence of model launches, because the catalogue is only useful if it stays current.

The measure of success will be the same as in year one: whether a creator, an agency or a developer can start and finish a project without leaving.

Three projects, start to finish

Numbers describe a platform. Three illustrative scenarios show what it is for.

A skincare launch, one afternoon. A two-person agency opens Chat Live with the client's brief and asks for a campaign concept. The concept comes back with a shot list; the agency renders the hero image on Seedream 5.0 in the generator, moves it to the photo editor to clean the label and extend the background, and sends the result to the video studio for a 10-second product loop on Seedance 2.5. FH Shorts cuts the loop into 9:16 and adds captions. Canvas lays out the still, the loop and the copy into a client presentation. Total tools opened outside FOTOhub: none. Total files downloaded before the client review: none.

A catalogue of 400 products, one week. An online shop connects its store through the WooCommerce plugin and pushes product photos to FOTOhub. The generator produces packshots and lifestyle scenes for every item; try-on renders the apparel on models; UGC Ads turns the twenty best sellers into short presenter videos. Everything lands back in the shop through the plugin, named by the shop's own convention, and the whole run is billed from one balance that the owner can read at any moment.

An agent that makes assets while writing code. A developer building a marketplace connects FOTOhub to Cursor over MCP. The agent generates placeholder-free product imagery as it scaffolds the pages, upscales the hero images, and files every URL into the seed data. When the marketplace goes live, the same developer switches to the Python SDK and the same key, and generation becomes part of the backend.

What a year taught us

Five things that were not obvious in September 2025 and are obvious now.

The orchestration layer is the product. Users do not experience models. They experience whether the thing they asked for arrived, at the price they were quoted, in the format the next step needs. Every month spent on the layer between the models paid back more than any single model launch.

Billing has to be boring. The fastest way to lose a creator's trust is a surprise on the invoice. Showing the price before every operation, on every model, on every surface, cost real engineering time in the first half of the year. It is the reason the second half could ship so fast.

Own at least one model. Reselling alone leaves a platform hostage to its suppliers' pricing. IDA Q changed the cost structure of the most common operations and, with it, what the platform could afford to give away.

Developers arrive later, and stay longer. The API was an afterthought in the first quarter. By the third it had its own console, its own SDKs and its own customers, because the people who build on FOTOhub bring their users with them.

Ship monthly, announce honestly. Every announcement on this blog names what shipped and what did not. The MCP article last week listed three gaps in its own feature. Readers noticed, and the trust that buys is worth more than a cleaner headline.

A short glossary

Credits. The single unit of usage across all 226 models. Every operation shows its price in credits before it runs. Plan credits renew every 30 days.

Buster Packs. One-time credit packs that stack on any plan, from 200 credits for 7 days to 50,000 credits with no expiry.

Gabriel. FOTOhub's orchestrator. Ask for a result and it picks the model, sets the parameters and routes the job. Available in the app, the API and MCP.

IDA. The family of FOTOhub's own models: IDA Q for image generation, IDA Voice for speech, IDA Y for the wider audio line.

Flows. Automations that chain models into a pipeline that runs without a person in the loop.

Teams. A plan for 3 to 50 people with a shared credit pool, roles and per-member limits enforced on the server.

MCP. The Model Context Protocol, the open standard that lets AI assistants call external tools. FOTOhub is both an MCP server (28 tools) and, in Chat Live, an MCP client.

Frequently asked questions

What is FOTOhub? A Creative AI hub: one subscription that gives access to 226 AI models for image, video, audio, chat, text, 3D and try-on, the studios built on top of them, and a developer API, all from one account and one balance.

How many people use it? More than 1 million users in 46 countries as of September 2026.

When was FOTOhub founded? September 11, 2025, in Poland, by Mateusz Ulewicz. The company is incorporated in Delaware.

How much has FOTOhub raised? $8.6 million to date, with Amazon Web Services and DGE3 Investment among its backers. The company was valued at $180 million in Q2 2026.

Does FOTOhub have its own models? Yes. IDA Q 1.0 for image generation launched in July 2026, and the IDA Voice and IDA Y family covers speech, music and sound.

Is there an API? Yes. 314 REST endpoints, Python and TypeScript SDKs (`pip install fotohub`, `npm install fotohub`), a CLI, an n8n node, webhooks and an MCP server with 28 tools. Documentation at docs.fotohub.app.

How much does it cost? Plans start at $8 a month (33 zł) for 300 credits. The free plan includes 50 credits and needs no card. Current prices at fotohub.app/public_prices.

Where does FOTOhub run? On Amazon Web Services, with its own GPU fleet in Frankfurt and Ireland and storage available across 16 regions.

A note from the founder

"A year ago we had one generator and a hypothesis," says Mateusz Ulewicz. "The hypothesis was that creators do not want ten AI tools, they want one place where all of it works together and one bill they can predict. Every month since then has been a test of that hypothesis, and every month the answer has been the same. People came for the models and stayed for the fact that the models talk to each other."

On what the second year is for: "The first year was about breadth. We had to prove that one platform could carry image, video, audio, chat and 3D at production quality, with billing that never lies and infrastructure that never surprises. We did that. The second year is about depth: our own models carrying more of the load, agents that run whole workflows, and a developer platform that is as good to build on as the app is to use. A million users is a milestone. It is not the goal."

On building from Poland for a global market: "We hire engineers who learned their craft shipping under pressure, and we let them ship. That culture, plus an American company structure and a cloud partner with sixteen regions, turns out to be a good recipe. Our users in 46 countries do not care where the servers are. They care that the render arrives."

Press materials, logos and company facts are at fotohub.app/press-kit. Investor information is at fotohub.app/investor-relations. Live platform status is at status.fotohub.app.

Start with 50 free credits at fotohub.app/generate/new, read the developer documentation at docs.fotohub.app, or see the full catalogue at fotohub.app/product/our-ai-models.