Deep Dream Generator

The 2015 DeepDream site, now a multi-model AI art community

Image Generation Freemium
Researched · Published · Reviewed
RECATOOLS Score
4.6 / 10
Capability
4
Value for money
5
Ease of use
6
ASEAN readiness
6
API quality
n/a
Founded
2015
HQ
Sofia, Bulgaria
Users
3m+ users
Launched
2015
Developer
Aifnet Ltd.

Overview

Deep Dream Generator, run by Sofia-based Aifnet Ltd., began in 2015 as a web front end for Google's DeepDream and is now a community AI art platform offering third-party image models such as Nano Banana 2, Flux 2 and Ideogram, video models including Veo 3.1 and Kling, and music generation, all paid for in "energy" credits.

Advertisement

Pricing

Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.

Free
$0/mo
Rechargeable energy by community level
  • 20–70 max energy, 1–4 an hour recharge
  • 3 PRO-model uses a day
  • 1–2 GB storage
  • No video or commercial use
Basic
$9/mo
750 energy a month
  • About 150 images, 10 videos or 25 music tracks
  • Unlimited Z-Image Turbo
  • PRO features and commercial use
  • 10 GB storage
Advanced
$19/mo
2,000 energy a month
  • About 500 images, 30 videos or 50 tracks
  • Unlimited Z-Image Turbo
  • PRO features and commercial use
  • 20 GB storage
Ultra
$99/mo
15,000 energy a month
  • About 3,750 images, 250 videos or 500 tracks
  • Unlimited Z-Image Turbo
  • PRO features and commercial use
  • 200 GB storage
Energy packs
$3.99–$599 one-time
120 to 50,000 energy
  • No monthly commitment
  • Commercial use for work made with pack energy
  • 5,000 energy for $79

Use cases

Creating psychedelic visual art with a distinctive neural network aesthetic Transforming landscape photography into surreal dreamlike compositions Historical AI art experiments using the original Deep Dream algorithm

What you can produce with Deep Dream Generator

  • Images from 22 models, including Nano Banana 2, Flux 2, Ideogram v4 and DaVinci2
  • Short AI videos with Veo 3.1, Kling 2.6, Seedance or MiniMax on paid plans
  • Music generation, image and video upscaling, and image editing and cut-outs
  • A public gallery with visible prompts, challenges and a community chat
  • A free start on rechargeable energy, then a plan or a one-time energy pack
  • Commercial use of outputs made on a paid plan or energy pack
Advertisement
RECATOOLS Verdict

Deep Dream Generator outlived the 2015 DeepDream wave by becoming a front end for other companies' models: Nano Banana 2, ChatGPT 2, Flux 2, Ideogram and Seedream for images, Veo 3.1 and Kling for video, alongside its own DaVinci2 and the legacy Deep Style. What it adds is the community: a public gallery with visible prompts, challenges and a members' chat, which makes it a good place for hobbyists to learn and share.

Everything runs on energy credits. The free pool is small and refills slowly, video needs a paid plan, and commercial use requires a plan or energy pack. Plans run $9 to $99 a month with one-time packs from $3.99, and there is no public API.

Desk research by RECATOOLS from public sources, written with AI assistance — not a hands-on test. How we review

What we found

Deep Dream Generator launched in 2015 as a web front end for Google's DeepDream algorithm and is now a multi-model image, video and music platform. It is operated by Aifnet Ltd. of Sofia, Bulgaria, and the about page credits Kaloyan Chernev as its creator. The release notes list four versions, 5.19 to 5.22, in September 2026 alone.

The image generator now leans mostly on other companies' models, not on a Stable Diffusion base as this page earlier said. A logged-out visitor can pick from 22, including Nano Banana 2, ChatGPT 2, Flux 2, SeeDream, Ideogram v4, Grok and Qwen, alongside DaVinci2, which the site calls SDXL-based, older in-house models such as Photonic, and the legacy Deep Style. Video runs through Veo 3.1, Kling 2.6, Seedance and MiniMax, with upscaling, editing and music tools alongside.

Everything costs "energy". Free accounts get a small rechargeable pool: 20 energy refilling at one an hour for a newcomer, up to 70 at four an hour for the top community level, with three PRO-model uses a day. A DaVinci2 image costs 4 energy, a Nano Banana 2 image 50 and a four-second Veo 3.1 clip 280. Paid plans run from $9 to $99 a month, each with PRO models, video, commercial-use rights and 10 to 200 GB of storage, and one-time energy packs start at $3.99. The FAQ says images carry no watermark and limits commercial use to work made on a paid plan or pack.

The community is what sets it apart: a public gallery with visible prompts, challenges, a "try this prompt" button and, since September 2026, a chat room where subscribers and top-level members can post. Trustpilot showed 4.3 from 79 reviews on 12 October 2026, with the community the commonest theme (19 of the 63 visible), though nearly all were posted in 2025, after the profile was claimed that January.

It suits hobbyists who enjoy sharing work. There is no public API, so developers and production teams are better served by the model providers directly.

Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.

Notable facts

  • Deep Dream was created by Google engineer Alexander Mordvintsev in 2015 as a tool to understand what features neural networks were detecting — the psychedelic visual output was an unexpected byproduct.
  • The Deep Dream algorithm was the first neural network visualization technique to generate viral public interest in AI-created art, predating the current AI art wave by 7 years.
  • In 2016, a Deep Dream-generated artwork sold at Christie's auction for $8,000 — one of the first AI artworks sold at a major auction house.

Frequently asked questions

Yes, with limited daily credits. Advanced plans provide more.
Psychedelic patterns with amplified features — dog faces, eyes, geometric structures — that emerge from neural network feature detection amplified to visible levels.
Yes. Upload any image and apply Deep Dream effects to it.
It uses the same algorithm as Google's original Deep Dream research, made accessible via a web interface.
The primary function transforms existing images. Basic text-to-image is also available.

About this listing

Researched on
Published on
Last reviewed

This entry was compiled from publicly available data including Deep Dream Generator's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Deep Dream Generator unless explicitly stated.

Data accuracy

Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.

For the latest details, please refer to Deep Dream Generator directly →

Spotted something out of date? Suggest an update →

Advertisement