# t5xxl_fp16.safetensors

> T5 v1.1 XXL text encoder (fp16) · T5 text encoder for FLUX.1 and SD3

t5xxl_fp16.safetensors is Google’s T5 v1.1 XXL text encoder in fp16, packaged for ComfyUI. FLUX.1 and SD3 read your prompt through it alongside CLIP. Put it in ComfyUI/models/text_encoders and load it in DualCLIPLoader (FLUX) or TripleCLIPLoader (SD3).

The Hugging Face repo needs no sign-in. If your RAM is under 32 GB, the fp8 version is half the size.

## Facts

- **Type:** Text encoder
- **Format:** Safetensors · fp16
- **Size:** 9.79 GB (9,787,841,024 bytes)
- **SHA256:** 6e480b09fae049a72d2a8c5fbccb8d3e92febeb233bbe9dfe7256958a9167635
- **ComfyUI folder:** ComfyUI/models/text_encoders/
- **Loader node:** DualCLIPLoader
- **License:** Apache-2.0

## What is t5xxl_fp16.safetensors?

T5 v1.1 XXL is the encoder half of Google’s T5 language model. FLUX.1 and Stable Diffusion 3 use it next to CLIP because it understands long, detailed prompts and spelled-out text far better than CLIP alone.

This file holds only the encoder, in fp16, as a single safetensors file that ComfyUI can load directly. The same file works for FLUX.1 dev and schnell, FLUX fine-tunes, SD3 and SD3.5.

## Download and install

- [Hugging Face · comfyanonymous/flux_text_encoders](https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/t5xxl_fp16.safetensors) · saves as `t5xxl_fp16.safetensors` · ComfyUI’s own repackage. Recommended.

```sh
cd ComfyUI/models/text_encoders
wget -O t5xxl_fp16.safetensors \
  https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/t5xxl_fp16.safetensors
sha256sum t5xxl_fp16.safetensors
# 6e480b09fae049a72d2a8c5fbccb8d3e92febeb233bbe9dfe7256958a9167635
```

Safetensors files hold tensors only and cannot run code. Still compare the SHA256: a failed download often leaves an HTML error page saved under the model’s filename.

## How to load it in ComfyUI

FLUX.1 loads T5 next to clip_l in DualCLIPLoader (type flux). SD3 loads it in TripleCLIPLoader with clip_g and clip_l. Older video models such as CogVideoX load it alone in Load CLIP.

- **Order does not matter:** DualCLIPLoader accepts T5 in either slot; the type setting decides how each is used.
- **Low on RAM?:** t5xxl_fp8_e4m3fn is half the size of fp16 with a small quality cost.

## Files a FLUX.1 workflow needs

FLUX.1 dev is a gated download: accept the license on Hugging Face first.

- [`flux1-dev.safetensors`](https://huggingface.co/black-forest-labs/FLUX.1-dev) · Diffusion model · `models/diffusion_models/`
- [`clip_l.safetensors`](https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/clip_l.safetensors) · Text encoder (CLIP-L) · `models/text_encoders/` · 246 MB
- [`ae.safetensors`](https://huggingface.co/black-forest-labs/FLUX.1-dev) · VAE · `models/vae/`

## Workflows: Uses this exact file

These workflows already load t5xxl_fp16.safetensors. Open one and run it as is.

- [VOID Video Inpainting ComfyUI | Smart Object Removal](https://www.runcomfy.com/comfyui-workflows/void-video-inpainting-comfyui-temporal-object-clean-up-workflow)
- [FLUX Redux ACE Face Swap | Precise Face Generator](https://www.runcomfy.com/comfyui-workflows/flux-redux-ace-face-swap-in-comfyui-photo-identity-transfer)
- [Create Coherent Scenes | Consistent Story Art Generator](https://www.runcomfy.com/comfyui-workflows/create-coherent-scenes-qwen-image-edit-wan-2-2-in-comfyui-cinematic-coherence-workflow)
- [Flux Krea Dev | Natural Text to Image](https://www.runcomfy.com/comfyui-workflows/flux-krea-dev-comfyui-workflow-natural-text-to-image-ai)
- [FLUX Kontext LoRA | Style Transfer](https://www.runcomfy.com/comfyui-workflows/comfyui-flux-kontext-lora-workflow-style-transfer-visual-remix)
- [Consistent Face 3x3 Generator](https://www.runcomfy.com/comfyui-workflows/consistent-face-3x3-generator)
- [ACE++ Face Swap ｜ Image Editing](https://www.runcomfy.com/comfyui-workflows/ace-plus-plus-face-swap)
- [InfiniteYou | Identity-Preserving Face Generation](https://www.runcomfy.com/comfyui-workflows/comfyui-infiniteyou-identity-preserving-face-generation-toolkit)
- [ACE++ Character Consistency](https://www.runcomfy.com/comfyui-workflows/ace-plus-plus-character-consistency)
- [AP Workflow 12.0 | Ready-to-Use Complete AI Media Suite](https://www.runcomfy.com/comfyui-workflows/ap-workflow-for-comfyui-ready-to-use-online-solution)
- [FLUX LoRA (RealismLoRA) | Photorealistic Images](https://www.runcomfy.com/comfyui-workflows/comfyui-flux-realismlora-workflow-photorealistic-ai-images)
- [CogvideoX Fun | Video-to-Video Model](https://www.runcomfy.com/comfyui-workflows/cogvideox-fun-video-to-video-model)
- [FLUX ControlNet Depth-V3 & Canny-V3](https://www.runcomfy.com/comfyui-workflows/comfyui-flux-controlnet-depth-and-canny)
- [Stable Diffusion 3.5](https://www.runcomfy.com/comfyui-workflows/stable-diffusion-3.5-models-in-comfyui)
- [FLUX IPAdapter V2 | XLabs](https://www.runcomfy.com/comfyui-workflows/xlabs-flux-ipadapter-v2)
- [Pyramid Flow | Video Generation](https://www.runcomfy.com/comfyui-workflows/pyramid-flow-video-generation)
- [FLUX Controlnet Inpainting](https://www.runcomfy.com/comfyui-workflows/flux-controlnet-inpainting-image-repair)
- [Stable Diffusion 3.5 vs FLUX.1](https://www.runcomfy.com/comfyui-workflows/stable-diffusion-3.5-vs-flux1)

## Workflows: Uses another version of this file

These workflows load a different precision or format of the same model. Select t5xxl_fp16.safetensors in their DualCLIPLoader node to use this version.

- [Wan 2.2 Video Restyle | First Frame Restyle for Consistent and Cinematic Video Generation](https://www.runcomfy.com/comfyui-workflows/wan2-2-first-frame-restyle-comfyui-video-generation)
- [Flux Kontext Zoom Out ComfyUI Workflow | Seamless Outpainting](https://www.runcomfy.com/comfyui-workflows/flux-kontext-zoom-out-in-comfyui-clean-canvas-expansion)
- [Omni Kontext | Seamless Scene Integration](https://www.runcomfy.com/comfyui-workflows/comfyui-omni-kontext-workflow-identity-aware-scene-composition)
- [FLUX Kontext Face Swap | Seamless Face Replacement](https://www.runcomfy.com/comfyui-workflows/comfyui-flux-kontext-face-swap-workflow-photoreal-swaps-localized-refinement)
- [FLUX Kontext Dev | Intelligent Image Editing](https://www.runcomfy.com/comfyui-workflows/flux-kontext-dev-comfyui-workflow-ai-image-editing-tool)
- [Put It Here Kontext | Object Replacement](https://www.runcomfy.com/comfyui-workflows/put-it-here-kontext-ai-object-replacement-comfyui-image-editing)
- [HiDream E1.1 | AI Image Editing](https://www.runcomfy.com/comfyui-workflows/hidream-e11-comfyui-workflow-ai-image-editing-model)
- [FLUX Kontext Preset | Scene Control](https://www.runcomfy.com/comfyui-workflows/flux-kontext-preset-comfyui-workflow-ai-scene-control)
- [FLUX Kontext OmniConsistency LoRA](https://www.runcomfy.com/comfyui-workflows/flux-kontext-omniconsistency-lora)
- [Flux Kontext 360 Degree LoRA](https://www.runcomfy.com/comfyui-workflows/comfyui-flux-kontext-360-degree-lora-workflow)
- [Flux Kontext Character Turnaround Sheet LoRA](https://www.runcomfy.com/comfyui-workflows/flux-kontext-character-turnaround-sheet-lora)
- [DreamO | Unified Multi-Task Image Customization Framework](https://www.runcomfy.com/comfyui-workflows/dreamo-unified-image-customization-framework)
- [FLUX Dev ControlNet | Multi-Condition ControlNet](https://www.runcomfy.com/comfyui-workflows/flux-dev-controlnet-union-pro-multi-condition)
- [ICEdit | Fast AI Image Editing with Nunchaku](https://www.runcomfy.com/comfyui-workflows/icedit-comfyui-workflow-efficient-ai-image-editing-with-nunchaku)
- [Flux TTP Upscale | 4K Face Restore](https://www.runcomfy.com/comfyui-workflows/flux-upscaler-ttp-face-restore-4k-upscaling)
- [PuLID Flux II | Consistent Character Generation](https://www.runcomfy.com/comfyui-workflows/pulid-flux-ii-in-comfyui-consistent-character-ai-generation)
- [Flux PuLID for Face Swapping](https://www.runcomfy.com/comfyui-workflows/realistic-face-swapping-with-flux-pulid)
- [FLUX | A New Art Image Generation](https://www.runcomfy.com/comfyui-workflows/comfyui-flux-a-new-art-image-generation)
- [Flux Consistent Characters | Input Text](https://www.runcomfy.com/comfyui-workflows/consistent-characters-with-flux-comfyui-workflow)
- [FLUX Inpainting | Seamless Image Editing](https://www.runcomfy.com/comfyui-workflows/comfyui-flux-inpainting-workflow)
- [FLUX Img2Img | Merge Visuals and Prompts](https://www.runcomfy.com/comfyui-workflows/comfyui-flux-img2img-workflow)
- [FLUX IPAdapter V1 | XLabs](https://www.runcomfy.com/comfyui-workflows/comfyui-flux-ipadapter)
- [CogVideoX-5B | Advanced Text-to-Video Model](https://www.runcomfy.com/comfyui-workflows/cogvideox5b-text-to-video-model-available-in-comfyui)
- [Stable Diffusion 3 (SD3) | Text to Image](https://www.runcomfy.com/comfyui-workflows/leverage-stable-diffusion-3-for-advanced-visuals)

## Same file, other names, and alternatives

- `t5xxl_fp16.safetensors` · This page · fp16 · 9.79 GB · Best prompt fidelity
- `t5xxl_fp8_e4m3fn.safetensors` · Same encoder · fp8 e4m3fn · 4.89 GB · Half the RAM and VRAM
- `t5xxl_fp8_e4m3fn_scaled.safetensors` · Same encoder · fp8 scaled · 5.16 GB · fp8 with better accuracy

## Common errors and fixes

### Value not in list: clip_name2: 't5xxl_fp16.safetensors' not in [...]

Why: The workflow asks for this exact filename, but ComfyUI cannot find it in models/text_encoders. Usually the file is missing, sits in a different folder.

Fix: Move the file into ComfyUI/models/text_encoders/, rename it to match the workflow (or pick it again in the dropdown), then refresh the node list or restart ComfyUI.

### SafetensorError: Error while deserializing header

Why: The download is incomplete, or the browser saved an HTML sign-in or error page under the model’s filename.

Fix: Check that the file is exactly 9,787,841,024 bytes and that its SHA256 matches the value above. Download it again from Hugging Face · comfyanonymous/flux_text_encoders if not.

## FAQ

### fp16 or fp8 for FLUX?

fp16 if you have about 32 GB of system RAM or more. Otherwise use `t5xxl_fp8_e4m3fn.safetensors`; most people cannot tell the difference in final images.

### Is there another place to download t5xxl_fp16.safetensors besides Hugging Face?

Hugging Face (comfyanonymous/flux_text_encoders) is the official source. Copies elsewhere are fine if the SHA256 matches `6e480b09fae049a72d2a8c5fbccb8d3e92febeb233bbe9dfe7256958a9167635`.

### Where do I put t5xxl_fp16.safetensors in ComfyUI?

In `ComfyUI/models/text_encoders/`. Then refresh the node list or restart ComfyUI so DualCLIPLoader lists the file.

### Which ComfyUI node loads t5xxl_fp16.safetensors?

[DualCLIPLoader](/comfyui-nodes/ComfyUI/dual-clip-loader). It is built into ComfyUI, so no custom nodes are needed.

### Can I use t5xxl_fp16.safetensors commercially?

Yes. T5 v1.1 XXL and this ComfyUI repackage are released under Apache-2.0.

## Sources

- [comfyanonymous/flux_text_encoders](https://huggingface.co/comfyanonymous/flux_text_encoders)
- [google/t5-v1_1-xxl](https://huggingface.co/google/t5-v1_1-xxl)

Canonical page: https://www.runcomfy.com/comfyui-workflows/model-files/t5xxl_fp16.safetensors
