Fal API Flux Pro V1.1

FLUX1.1 [pro] is Black Forest Labs' refresh of the original Pro checkpoint, and their own launch pitch for it was blunt: several times faster generation than the first Pro model, with better prompt adherence and more variety in the results. This node is fal's wrapper around that newer endpoint. If you're choosing between "Fal API Flux Pro" and "Fal API Flux Pro V1.1" in this pack's node list, V1.1 is the one to default to - it's the same undistilled, API-only tier, just the version BFL currently wants you using.

The README lists it plainly as "an updated node for text-to-image generation using FLUX.1 [pro] V1.1," and the schema backs that up: it's a drop-in replacement for the plain Pro node, not a different kind of tool. Same inputs, same output, same billing model - every call still goes out over the network to fal.ai and comes back as a finished IMAGE, not a local render.

Inputs. prompt (multiline - write full sentences, Flux's T5 encoder parses natural language rather than tag soup), width and height (1024 default, step 8, Flux's native resolution), num_inference_steps (28 default, 1–100), guidance_scale (3.5 default, 0.1–40), num_images (1–4 per call), enable_safety_checker, and safety_tolerance - a 1–6 enum fal documents as running from strictest to most permissive content filtering. Treat that last one as informational rather than load-bearing: BFL has previously yanked this exact parameter from their Kontext Pro/Max API without any announcement, so don't be surprised if it eventually stops mattering here too. seed is optional and defaults to 0 for a fresh draw each run.

Because it's the Pro line, there's no LoRA slot and never will be - Pro's weights aren't published for anyone to fine-tune, on fal's servers or anywhere else. If your workflow leans on a custom LoRA, one of the Dev-with-LoRA nodes in this pack is the right tool, not this one.

Installing it. Same as every node in this pack: ComfyUI Manager (search "ComfyUI-Fal-API-Flux") is the path of least resistance. Manual installers should note the README's git clone line points at a placeholder your-username repo rather than the real yhayano-ponotech one - a leftover copy-paste bug, easy to miss, easy to fix by using the correct URL. After cloning, pip install -r requirements.txt, then the step that actually matters: a config.ini file in the pack's own folder, [falai] section, api_key = .... Restart ComfyUI once that's in place.

Troubleshooting. This node has essentially zero measured traffic on comfy.icu, which tracks with it being a lesser-known variant of a lesser-known pack - so if you hit a wall, don't assume it's you: check the same short list first. Missing or malformed config.ini, a fal account that isn't provisioned for Pro-tier Flux specifically, and the ComfyUI console for fal's actual error text rather than whatever the node surfaces on its own. If your fal account works fine on the plain Pro node but not here, it's worth double-checking your account has V1.1 access specifically - fal sometimes gates newer model endpoints separately from older ones.