← All companies & organizations

Neo-labs / Company profile

Black Forest Labs

What Black Forest Labs does

Black Forest Labs (BFL) is a frontier visual AI research lab building multimodal “world” foundation models aimed at creative and real-world applications. The company’s models and infrastructure center on the FLUX family: text-to-image generation and in-context image editing (FLUX.1 / FLUX.2 / FLUX.2 [klein]) plus an expanding multimodal roadmap that adds video, audio, and action-oriented capabilities under FLUX 3. BFL distributes FLUX through a production-oriented API and developer tooling (including asynchronous generation endpoints and a “FLUX Tools” suite), while also offering open-weight access through licensed distributions for users who want to run models on their own infrastructure.

BFL’s business model is a combination of (1) paid API access to higher-tier models and (2) enterprise and platform integrations (for teams deploying at production scale with reliability, security, and workflow fit). For individual developers and technical audiences, the company also supports experimentation via a web playground and publicly accessible/open-weight routes, plus integrations with third-party platforms such as Replicate and others (notably for FLUX.1 tiers). Strategically, BFL positions itself as both an open research lab and an enterprise-grade deployment partner—pairing published/open model artifacts with proprietary endpoints, safety mitigations, and partner deployments across creative and commercial workflows.

News

Company record

Aug 04, 2026 · Official · Black Forest LabsFLUX 3 Video, Part 1: Generation is generally available

Black Forest Labs announced generally available release of an initial FLUX 3 Video generation capability via the BFL API and select partners, including clips up to 20 seconds with native audio.

↗
Jul 23, 2026 · Official · Black Forest LabsFLUX 3 x mimic: video-action models running on robots

Black Forest Labs described FLUX 3 running on robots via an early-access collaboration with mimic, introducing FLUX-mimic as a video-action model built from the FLUX 3 backbone and validated in real-world robot deployment testing (described as involving Audi).

↗
Jul 23, 2026 · Official · Black Forest LabsFLUX 3: multimodal foundation model with video, audio and action-oriented roadmap

Black Forest Labs announced FLUX 3 as a multimodal foundation model, describing early access availability and detailing joint learning across images, videos, and audio under a unified architecture, plus an action prediction roadmap.

↗
Jun 18, 2026 · Official · Black Forest LabsOur co-founder and CEO urges G7 leaders to back open innovation

Robin Rombach’s G7 discussion centered on open innovation and responsible development in visual AI, and included an overview of BFL’s mission and framing for visual world models.

↗
Dec 01, 2025 · Official · Black Forest LabsLaying the Foundations for Visual Intelligence—Our $300M Series B

The company announced a $300M Series B at a $3.25B post-money valuation, naming co-leads and describing partner and investor participation.

↗
Aug 01, 2024 · Official · Black Forest LabsAnnouncing Black Forest Labs; release of FLUX.1 suite and $31M Series Seed

Black Forest Labs publicly announced its launch, released the FLUX.1 suite of text-to-image models in three variants, and reported closing a $31M Series Seed round led by Andreessen Horowitz.

↗
Source map · 3 recurring channels · 65 references
Black Forest LabsOfficial Site · monitor↗Axel Sauer on XPrincipal X · monitor↗MTS Frontier / Control RoomCurated Feed · monitor↗Black Forest LabsOfficial Reference · reference↗Official company sourceOfficial Reference · reference↗sec.govRegulatory · reference↗13f.infoReporting · reference↗a16z.comReporting · reference↗aiwiki.aiReporting · reference↗app.dealroom.coReporting · reference↗arxiv.orgReporting · reference↗assets.ctfassets.netReporting · reference↗assets.kpmg.comReporting · reference↗axios.comReporting · reference↗bf-core-prod-public-documents.s3-eu-west-1.amazonaws.comReporting · reference↗bfl.aiReporting · reference↗bizprofile.netReporting · reference↗caplight.comReporting · reference↗cbinsights.comReporting · reference↗crunchbase.comReporting · reference↗d18rn0p25nwr6d.cloudfront.netReporting · reference↗datocms-assets.comReporting · reference↗de.linkedin.comReporting · reference↗de.wikipedia.orgReporting · reference↗dealroom.coReporting · reference↗docs.eu.bfl.aiReporting · reference↗en.wikipedia.orgReporting · reference↗engine-ventures.files.svdcdn.comReporting · reference↗fcf.deReporting · reference↗files.nitrd.govReporting · reference↗forbes.comReporting · reference↗forgeglobal.comReporting · reference↗formdexplorer.comReporting · reference↗forms13f.comReporting · reference↗frontrunner.fyiReporting · reference↗fundraisingfox.comReporting · reference↗gen-ai-mastery-book-v1.vercel.appReporting · reference↗github.comReporting · reference↗google.comReporting · reference↗gtai.deReporting · reference↗handelsblatt.comReporting · reference↗heise.deReporting · reference↗help.bfl.aiReporting · reference↗insiderllm.comReporting · reference↗intro-act.comReporting · reference↗investor.bgmsglobal.comReporting · reference↗ir.bioagelabs.comReporting · reference↗labscompanies.comReporting · reference↗linkedin.comReporting · reference↗marketscreener.comReporting · reference↗news.bloomberglaw.comReporting · reference↗notioncapital.comReporting · reference↗openreview.netReporting · reference↗pitchbook.brightspotcdn.comReporting · reference↗quantlogix.aiReporting · reference↗reddit.comReporting · reference↗researchgate.netReporting · reference↗rothschildandco.comReporting · reference↗rss.globenewswire.comReporting · reference↗s2.q4cdn.comReporting · reference↗s24.q4cdn.comReporting · reference↗s3-us-west-2.amazonaws.comReporting · reference↗sacra.comReporting · reference↗seedtable.comReporting · reference↗siliconvalleyinvestclub.comReporting · reference↗techcrunch.comReporting · reference↗wbaf2019.istanbulReporting · reference↗zh.wikipedia.orgReporting · reference↗

Still resolving: Newsroom · Official X

Funding

Latest disclosed valuation$3.3B
Tracked capital$457.7M4 sourced rounds

What it builds

FLUX 3

A multimodal foundation model intended to generate and predict across image, video, audio, and action-related capabilities, with staged rollouts including early access and subsequent general availability for FLUX 3 Video features.

source ↗
FLUX 3 Video

Video generation capabilities under FLUX 3, released generally available in an initial form (including up to 20-second clips with native audio) via the BFL API and select partners.

source ↗
FLUX API (bfl.ai / docs.bfl.ai)

BFL’s production API for asynchronous media generation with endpoints for multiple FLUX variants, including global and regional routing and documented generation request/polling behavior.

source ↗

Milestones & partnerships