A first native SD1.5 text-to-image workflow, from checkpoint to saved image

File and download boundary: ComfyUI World provides source references and metadata only. Model and LoRA weights, node packages, and workflow files stay with their original providers; this site does not store, mirror, or proxy them. Downloads and access conditions are controlled by each provider.

Purpose and scope

Use the official baseline to learn node roles and establish a working environment without installing a large custom-node collection. This is an instructional path, not a generated example tested by this site.

Source-supported context

The basic example connects checkpoint, positive/negative text, an initial latent, KSampler, VAE Decode and a save endpoint. The source link is retained rather than presenting an invented JSON as the downloaded official graph. Source 1 Source 2

Before you start

1. Select the SD1.5 checkpoint specified by the example and confirm required MODEL, CLIP and VAE roles.

2. Start a core installation with permissions to read models and write outputs.

3. Use the official workflow/resource entry; an arbitrary newer-family checkpoint is not a substitute.

Steps and checks

The sequence below is an editorial procedure based on the cited context, not a diagnosis already confirmed for your environment.

Step 1. Obtain the original example or current basic template and check that its core node classes are available.

Step 2. Select the real checkpoint, encode positive/negative conditions through CLIP and choose modest Empty Latent dimensions/batch size.

Step 3. Check that KSampler uses the matching model/conditions, then decode with the paired VAE and connect a save endpoint.

Step 4. Run one small image and record seed, settings, artifact identity, versions and result, diagnosing load, sample, decode and save stages separately.

Completion check

A valid saved result exists and every object's origin/destination is explainable from reproducible settings.

Limitations and cautions

This pack contains neither the model weights nor the original upstream JSON bytes. upstreamUrl is the author's entry point, not a local mirror; no GPU-generated sample is claimed.

Original sources

• Text-to-image workflow — original record checked 2026-09-21.
• ComfyUI model troubleshooting — original record checked 2026-09-21.

Localization prepared: 2026-09-22. The source-check dates above were inherited from the existing article; this translation does not claim they were all rechecked today. No GPU run, minimum-memory measurement or runtime guarantee is supplied.

Related troubleshooting and guides

• Value not in list: why a downloaded model does not appear in ComfyUI
• CLIP input is invalid: None — trace where the text encoder should come from
• CUDA out of memory: separate model loading, sampling and VAE decoding