fal released FLUX 3 Max on 19 October 2026, a video model its research team post-trained on Black Forest Labs’ FLUX 3, and priced it at a launch rate of $0.0425 per second of 720p video, down from a list price of $0.17. On fal’s own pages, Google’s Veo 3.1 lists $0.20 per second without audio, so the launch rate is about 79% lower, by our arithmetic.
The rank, speed and quality figures below are fal’s and Black Forest Labs’ own. All prices are as listed on fal’s model and pricing pages on 19 October 2026, and the per-clip and percentage figures are our arithmetic.
Business Pill · COST PER TASK
A short explainer on cost per task: paying for the finished job rather than the token. It teaches the general idea only and says nothing about any company in this story.
The key insight: As we read it, the price gap is widest during the launch window: at $0.0425 per second until 7 November, a 10-second FLUX 3 Max clip on fal costs about $0.43, against $2.00 for Veo 3.1 without audio, by our arithmetic.
fal wrote on X: “Introducing FLUX 3 Max, the latest post-trained video model from fal Research in partnership with @bfl_ai. FLUX 3 Max is tied for #1 in image-to-video quality on independent third-party benchmarks, generating a 10-second, 720/768p video in ~12 seconds.”
Black Forest Labs posted: “Tied for #1 in image-to-video. 22× faster than Seedance 2.5 at the same quality.” It added that “FLUX 3 is a model designed for customization. @fal adapted and optimized it to create FLUX 3 Max.”
fal’s page says Black Forest Labs “brought the foundation model; fal Research brought post-training, inference optimization and systems engineering.” fal’s page lists “Two at launch”, text to video and image to video, while both companies’ X posts also link a keyframes-to-video endpoint on fal.
Price of a 10-second video on fal at each model’s listed per-second rate, read on fal’s model and pricing pages on 19 October 2026: FLUX 3 Max at its launch rate (to 7 November) and list price, Google Veo 3.1 (without audio) and Veo 3.1 Fast, Kling Video v3 Pro, MiniMax H3 Max and H3 Max Turbo. Per-clip figures are our arithmetic.
What Was Posted on X
Introducing FLUX 3 Max, the latest post-trained video model from fal Research in partnership with @bfl_ai.
FLUX 3 Max is tied for #1 in image-to-video quality on independent third-party benchmarks, generating a 10-second, 720/768p video in ~12 seconds.
For the next three weeks, FLUX 3 Max is 75% off, making it the most affordable frontier video model at its quality level.
fal announced FLUX 3 Max at 18:34 UTC on 19 October, calling it a post-trained video model from fal Research in partnership with Black Forest Labs, tied for #1 in image-to-video quality and generating a 10-second video in about 12 seconds. The post gives the 75% discount as running for the next three weeks; the piece quotes it for the launch claims.
Introducing FLUX 3 Max.
Fastest in the west. Fast enough to try anything. And everything.
Generate a 10-second video in 12 seconds.
Tied for #1 in image-to-video. 22ร faster than Seedance 2.5 at the same quality.
FLUX 3 is a model designed for customization. @fal adapted and optimized it to create FLUX 3 Max.
Black Forest Labs, which built the FLUX 3 base model, posted at 18:51 UTC that FLUX 3 Max is tied for #1 in image-to-video and 22ร faster than Seedance 2.5 at the same quality, and that fal adapted and optimized FLUX 3 to create it. Its follow-up post describes the discount as 75% off for the next 2 weeks.
The Price and the Launch Rate
fal’s page lists FLUX 3 Max at “$0.0425 / second List $0.17, this rate ends November 7” for 720p, “billed per second of output video, with no minimums and no subscription.” That is 25% of the list price, the 75% discount both companies cite.
At those rates, by our arithmetic, a 10-second clip costs about $0.43 at launch and $1.70 at list. On fal’s pages, Veo 3.1 is charged “$0.20 without audio or $0.40 with audio for 720p or 1080p,” or $2.00 for 10 seconds without audio, and Veo 3.1 Fast is listed “From $0.10/s.”
fal’s pricing table lists Kling Video v3 Pro at $0.14 per second, MiniMax H3 Max at $0.05 and H3 Max Turbo at $0.025, so H3 Max Turbo is cheaper per second than the FLUX 3 Max launch rate. At its list price, FLUX 3 Max costs more per second than Kling Video v3 Pro and less than Veo 3.1.
The two companies give different lengths for the deal. fal’s post says “For the next three weeks,” and Black Forest Labs says “75% off for the next 2 weeks.” fal’s page gives the end date as 7 November, 19 days after launch.
The Benchmark Claim
fal’s page says FLUX 3 Max “scores 1,139 Elo in Artificial Analysis’s preliminary image-to-video arena, tied for #1 in blind human preference votes.” Its table lists MiniMax H3 (768p) at 1,141, Seedance 2.5 at 1,139 and Gemini Omni Flash 1.1 at 1,134.
fal notes that FLUX 3 Max is tied for #1 because “its confidence interval overlaps the leaders, with MiniMax H3 listed first on that board at 1,141.” The same table gives the time for a 10-second video as 12.45 seconds for FLUX 3 Max, 83.26 for MiniMax H3 and 277.80 for Seedance 2.5.
fal says post-training “lifted FLUX 3 by 50 Elo points, from 1,089 to 1,139, while inference work cut a 10-second video from 126.44 seconds to 12.45: about 10x faster.”
The German Lab Behind the Base Model
Black Forest Labs built FLUX 3, the base model. On the same day, deutschland.de, the portal run on behalf of Germany’s Federal Foreign Office, reported that Germany’s federal government plans to announce an AI partnership with Nvidia and that “Around a dozen German companies are expected to take part in the project,” including Siemens, Audi and Black Forest Labs, as FourWeekMBA reported.
The Structural Read
As we read it, the comparison fal draws is quality per second of waiting: its own table puts FLUX 3 Max level with Seedance 2.5 at 1,139 Elo while making a 10-second video in 12.45 seconds rather than 277.80.
As we read it, the discount is time-limited and the list price is what follows: at $0.17 per second FLUX 3 Max costs more than Kling Video v3 Pro at $0.14 on fal, and MiniMax H3 Max Turbo is cheaper per second even at the launch rate.
As we read it, the split of work is explicit in fal’s own description: Black Forest Labs supplied the base model and fal Research the post-training and inference work, with the model served on fal’s API.
fal, FLUX 3 Max page, 19 October 2026
“Black Forest Labs brought the foundation model; fal Research brought post-training, inference optimization and systems engineering.”
Three Implications
A LAUNCH RATE WITH AN END DATE fal lists $0.0425 per second until 7 November and $0.17 after; the two companies describe the deal as three weeks and two weeks.
A RANK SHARED, NOT HELD ALONE fal says FLUX 3 Max is tied for #1 in a preliminary Artificial Analysis snapshot, with MiniMax H3 listed first at 1,141 Elo.
A GERMAN BASE MODEL Black Forest Labs built FLUX 3, and deutschland.de lists it among German companies expected in Germany’s planned AI partnership with Nvidia.
The framework opens: “AI has crossed from the training era to the inference era. This is not a technical nuance. It is a fundamental economic restructuring that changes who makes money, how they make it, and where value concentrates in the AI economy”.
As we read it, FLUX 3 Max puts that shift into a single product: the base model came from Black Forest Labs, and fal’s pitch is the inference work on top, a 10-second clip in about 12 seconds sold by the second. The framework was published in March 2026, before FLUX 3 Max, and does not mention fal or Black Forest Labs.
What Is Not Established
The Elo scores and generation times are as published by fal, citing a preliminary Artificial Analysis snapshot. The publication could not read Artificial Analysis’s arena page directly, as it loads by script.
The price comparison uses the rates listed on fal only, with FLUX 3 Max at its launch rate unless stated, and Veo 3.1’s rate covers 720p or 1080p while FLUX 3 Max’s is for 720p. Prices on other platforms may differ, and the per-clip and percentage figures are our arithmetic.
Business Engineer Framework
The State of The Inference Economy
A Business Engineer analysis of how the shift from training to inference changes who makes money in AI, and how.
fal released FLUX 3 Max on 19 October 2026, post-trained on Black Forest Labs’ FLUX 3, and says it is tied for first in image-to-video quality while making a 10-second clip in about 12 seconds. Its launch rate of $0.0425 per second, until 7 November, is about 79% below Veo 3.1’s $0.20 on fal, by our arithmetic.
A note on sourcing. This piece is built on fal’s own FLUX 3 Max page and X posts and on Black Forest Labs’ X posts, all of 19 October 2026. Elo scores and generation times are fal’s published figures, citing a preliminary Artificial Analysis snapshot that the publication could not read directly. All prices are as listed on fal’s model and pricing pages on 19 October. Per-clip figures and percentages are our arithmetic. Nothing here is a forecast, and nothing here is financial or investment advice.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
