Flux
Black Forest Labs' image models — open weights plus hosted API
Overview
FLUX is Black Forest Labs' image-generation model family, built by ex-Stable Diffusion researchers and known for strong prompt adherence and in-image text rendering. FLUX.2 is the current line; schnell is free and Apache-2.0, higher tiers are commercial API models.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 27 Sep 2026 and may be out of date or incomplete.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.
- Self-host or run via API
- Commercial use allowed
- Fastest inference
- Open weights on Hugging Face
- Non-commercial license
- Strong ComfyUI/LoRA ecosystem
- Available via BFL API, fal.ai, Replicate
- Higher quality and speed
- Commercial licensing included
- Top-tier output quality
- Volume/enterprise agreements
- Dedicated support
Use cases
What you can produce with Flux
- Photorealistic product or lifestyle photography for e-commerce listings — generating studio-quality images of objects with accurate lighting and shadow at up to 4MP (FLUX1.1 Pro Ultra)
- Legible in-image typography — creating social media graphics, book covers, signage mockups, or advertising banners with accurately rendered text overlaid in the scene
- Consistent character portraits across a series — using FLUX.1 Kontext Pro with image-to-image conditioning to maintain the same face, outfit, or brand mascot across multiple shots without fine-tuning
- Custom brand-style image pipelines — training a LoRA on a company's visual identity and bulk-generating on-brand assets via the BFL API for marketing teams
- Contextual image editing without re-generating from scratch — swapping backgrounds, modifying objects, or changing clothing on an existing photo using the Kontext editing variants
- High-volume automated content generation — integrating the Schnell variant (Apache 2.0-licensed, ~$0.003/image) into a CMS or SaaS product to produce thumbnails, illustrations, or UI mockups at scale
- Artistic style transfer and portrait generation — producing pointillist illustrations, abstract emotional compositions, or professional-grade headshots from text descriptions for digital publishing and advertising
FLUX, from Black Forest Labs, remains one of the strongest image-generation families available — praised for prompt adherence, in-image text rendering, and quality that rivals or beats Midjourney and DALL·E on plenty of tasks. The company raised $300 million at a $3.25 billion valuation in a Series B backed by a16z, NVIDIA and Salesforce Ventures, and FLUX now runs inside production pipelines at companies including Adobe, Canva and Vercel. It's a model family, not a polished consumer app: access runs through the API, ComfyUI, or third-party front-ends, and licensing varies by tier — schnell is free and Apache-2.0, while dev and the pro/max API tiers carry commercial terms. Outstanding capability for teams willing to work at the model layer rather than a point-and-click studio.
What we found
Black Forest Labs closed a $300 million Series B in December 2025 at a $3.25 billion valuation, with a16z, NVIDIA, Salesforce Ventures, General Catalyst and Temasek all participating — a strong signal for a two-year-old company whose FLUX models now sit inside production pipelines at Adobe, Canva, Meta, Microsoft and Picsart — the ElevenLabs and Vercel names travelling with that list appear nowhere on Black Forest Labs' own site.
Reviewers and developers consistently rate FLUX's in-image text rendering — legible signs, labels, actual typography — well ahead of rival models, alongside strong hands, faces and photorealism that holds up against Midjourney. The Apache-2.0 schnell variant earned genuine praise as a free-for-commercial-use option, and that torch has passed to FLUX.2 klein's 4B model, released January 2026 — also Apache 2.0, ungated, and the lab's most downloaded release. Read klein carefully, though: only the 4B models are Apache, while the 9B versions sit under the non-commercial FLUX licence.
The gripes are structural rather than about quality: the first-party option is a Playground, not a consumer app, so casual users still mostly reach for a third-party front-end or ComfyUI; there are no pro or max weights to deploy, since those are API-only, while FLUX.2 dev is local-only with no hosted API; the non-commercial dev licence trips people up; and per-image API costs on hosted platforms add up fast at volume. LoRA and fine-tuning support still trails Stable Diffusion's older, larger ecosystem. Sentiment is decisively positive among developers and content teams, more mixed among solo creators who just want to type a prompt and get an image.
The entry is a generation behind. FLUX 3 is the current flagship and no longer only an image model: one multimodal model spanning image, video, audio and action prediction, with FLUX 3 Video shipping in August 2026 at up to twenty-second clips to 4K, and FLUX 3 Action in September as an open-weight 7B world-action model under a new community licence. Video is priced per second of output, six cents for a draft to eighty for 4K. One figure we could not verify is the 24GB VRAM requirement: the dev repository is gated, and the lab publishes only klein's 13GB.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
Notable facts
- The Black Forest Labs founding team created the original Stable Diffusion architecture at Ludwig Maximilian University Munich, then commercialised it at Stability AI before leaving to found BFL.
- FLUX.1 was released just 10 days after Black Forest Labs was founded, setting a record for fastest model release by a newly formed AI lab.
- The schnell (German for 'fast') variant generates high-quality images in just 1-4 steps, compared to 20-50 steps for earlier diffusion models.
Frequently asked questions
About this listing
This entry was compiled from publicly available data including Flux's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Flux unless explicitly stated.
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 Flux directly →
Spotted something out of date? Suggest an update →
Related reading
Alternatives to Flux
More in Image Generation
