- 🆕 Black Forest Labs' top-tier Flux 3 text-to-video generation endpoint
- 🎬 Turns a text prompt into clips with synchronized audio
- 📏 Up to 20 seconds, FHD 1920×1088, 24 fps per request
- 💬 Multilingual speech with strong lipsync, effects and ambience
- 🧠 Built on one multimodal model trained jointly across images, video, audio
- ⚡ Draft mode previews ideas cheaply before full-quality rendering
- 🔧 Keyframes, continuation and agentic chaining into multi-shot sequences
- 🏢 From Freiburg-based Black Forest Labs, founded by ex-Stability AI staff
Black Forest Labs is a generative AI company based in Freiburg im Breisgau, Germany, founded by former members of Stability AI. The lab is best known for developing the Flux family of text-to-image models, which generate images from natural language prompts and have quickly…
Explore 6 more models by Flux (text-to-image model) →Flux 3 Max is the highest-tier text-to-video configuration of Black Forest Labs' Flux 3 line, generating a video clip — with sound — directly from a written prompt. Per BFL's documentation, a single request can return up to 20 seconds at FHD resolution (1920 × 1088 for 16:9) at 24 fps, with multilingual speech, effects and ambience produced alongside the frames rather than dubbed in afterwards.
The jump from the previous generation is structural. The Flux 2 family on this catalog — Flux 2 Pro, Flux 2 Max and the editing-oriented Flux 2 Max Edit — is image-only. Flux 3 is described by BFL as a single multimodal model that mixes modalities and can generate images and video-with-audio jointly, from text alone or from image and video references; in preliminary midtraining evaluations the lab reports significant improvement over earlier Flux versions on complex prompts and text generation.
Within Flux 3, this entry is the Max-tier sibling of Flux 3 text-to-video and complements Flux 3 Max image-to-video, as well as Flux 3 First Last Frame and Flux 3 Extend. The same underlying model also powers Flux 3 image generation and Flux 3 editing.
Supporting features documented by BFL include keyframe control, video-to-video reference transfer, generative video-audio continuation, broad style range, and agentic chaining of clips into longer multi-shot sequences. A draft mode returns a fast, low-cost preview that can later be re-rendered at full quality.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
Daily usage · Flux 3 Max · Organic sellers
| Seller | Rep↓ | Sales | Video | Routing | $ / s | $ / video |
|---|---|---|---|---|---|---|
| ▲ Apex Ant 0x73b4…e736 | 81.71 | $1.09 | 38 s | #1 | $0.0288 | — |
"Best price" and the seller table are live Antseed catalog data (advertised $/1M tokens — or $ per generated image for unit-billed image models — not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what the VPR desktop app ships with): a trust ≥ 60 gate on the effective reputation, then cheapest-first among routable sellers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.