Skip to content

Releases: tngklp/ComfyUI-Prompt-Studio

Prompt Studio for ComfyUI v1.1.7

Choose a tag to compare

@tngklp tngklp released this 27 Sep 22:42

Prompt Studio for ComfyUI v1.1.7

ComfyUI extension release. Install through ComfyUI Manager or extract the ZIP into
custom_nodes.

Download

Download Prompt-Studio-ComfyUI-v1.1.7.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.

Install

  1. Extract the ZIP so that the folder ComfyUI-Prompt-Studio sits inside
    ComfyUI/custom_nodes.
  2. Restart ComfyUI.
  3. Open the floating Prompt Studio button, or use Extensions > Prompt Studio.
  4. Open Settings and choose a provider.

ComfyUI Manager is the simpler path and handles updates. Use the ZIP only for a manual
install.

What's new

  • Direct GGUF now works with llama-cpp-python 0.4.x. 1.1.6 widened one version
    check but missed a second, so the runtime was accepted in Settings and then refused
    per model with "Model setup is incomplete - the installed llama-cpp-python 0.4.0 does
    not support the qwen35 Direct adapter". That second check also blocked Gemma, so every
    Direct model failed. Both checks now share one supported range.
  • Anima no longer assumes every character is a girl. Selecting one female and one
    male character wrote 2girls. The gender of each character was being discarded before
    the prompt was built. A one-girl-one-boy selection now writes 1girl, 1boy.
  • The fix applies without deleting anything. The character catalogue records a
    format version, so an out-of-date cache is refreshed on the next launch rather than
    waiting up to 30 days.
  • Anima's quality tags remain masterpiece, best quality, highres, score_9.

Requirements

  • ComfyUI
  • Python 3.10 or newer
  • At least one configured provider

Models, API keys, and llama.cpp binaries are not bundled.

Looking for the version that runs without ComfyUI? See the
Standalone guide.

Prompt Studio for ComfyUI v1.1.6

Choose a tag to compare

@tngklp tngklp released this 27 Sep 22:30

Prompt Studio for ComfyUI v1.1.6

ComfyUI extension release. Install through ComfyUI Manager or extract the ZIP into
custom_nodes.

Download

Download Prompt-Studio-ComfyUI-v1.1.6.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.

Install

  1. Extract the ZIP so that the folder ComfyUI-Prompt-Studio sits inside
    ComfyUI/custom_nodes.
  2. Restart ComfyUI.
  3. Open the floating Prompt Studio button, or use Extensions > Prompt Studio.
  4. Open Settings and choose a provider.

ComfyUI Manager is the simpler path and handles updates. Use the ZIP only for a manual
install.

What's new

  • Direct GGUF no longer rejects a working runtime. llama-cpp-python 0.4.x was
    reported as "installed, but the runtime is not usable", which blocked Direct GGUF
    entirely. The supported range now runs from 0.3.34 up to, but not including, 0.5.0,
    so a CUDA cu130 build such as 0.4.0+cu130 is accepted. The 0.5.x series stays
    rejected because it has not been validated. Native compatibility and GPU execution are
    still exercised only when a Direct model actually loads.
  • Anima no longer assumes every character is a girl. Selecting one female and one
    male character wrote 2girls. The subject count tag is now stated explicitly and
    derived from each character's own gender, so a mixed pair is 1girl, 1boy.
  • Anima's quality tags are now masterpiece, best quality, highres, score_9.
    Previously masterpiece, best quality, score_7.

Requirements

  • ComfyUI
  • Python 3.10 or newer
  • At least one configured provider

Models, API keys, and llama.cpp binaries are not bundled.

Looking for the version that runs without ComfyUI? See the
Standalone guide.

Prompt Studio for ComfyUI v1.1.5

Choose a tag to compare

@tngklp tngklp released this 27 Sep 14:20

Prompt Studio for ComfyUI v1.1.5

ComfyUI extension release. Install through ComfyUI Manager or extract the ZIP into
custom_nodes.

Download

Download Prompt-Studio-ComfyUI-v1.1.5.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.

Install

  1. Extract the ZIP so that the folder ComfyUI-Prompt-Studio sits inside
    ComfyUI/custom_nodes.
  2. Restart ComfyUI.
  3. Open the floating Prompt Studio button, or use Extensions > Prompt Studio.
  4. Open Settings and choose a provider.

ComfyUI Manager is the simpler path and handles updates. Use the ZIP only for a manual
install.

What's new

  • Fixed "Generate prompt" for every image target. Qwen Image 2.1 (T2I and Edit),
    Krea 2 and Anima returned INVALID_DURATION: The selected mode does not accept a duration. Generating and refining a video prompt were never affected.
  • Image targets now ignore the duration value the studio sends with every request, since
    they have no duration of their own. Previously the request was rejected for carrying a
    field it could not use.
  • Only 1.1.5 changed how this works. 1.1.3 broke Refine and 1.1.4 broke Generate; if you
    are on either, update to 1.1.5.

Requirements

  • ComfyUI
  • Python 3.10 or newer
  • At least one configured provider

Models, API keys, and llama.cpp binaries are not bundled.

Looking for the version that runs without ComfyUI? See the
Standalone guide.

Prompt Studio for ComfyUI v1.1.4

Choose a tag to compare

@tngklp tngklp released this 27 Sep 14:07

Prompt Studio for ComfyUI v1.1.4

ComfyUI extension release. Install through ComfyUI Manager or extract the ZIP into
custom_nodes.

Download

Download Prompt-Studio-ComfyUI-v1.1.4.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.

Install

  1. Extract the ZIP so that the folder ComfyUI-Prompt-Studio sits inside
    ComfyUI/custom_nodes.
  2. Restart ComfyUI.
  3. Open the floating Prompt Studio button, or use Extensions > Prompt Studio.
  4. Open Settings and choose a provider.

ComfyUI Manager is the simpler path and handles updates. Use the ZIP only for a manual
install.

What's new

  • Fixed generating a prompt for an image target. The previous release fixed Refine for
    Anima, Qwen Image 2.1 and Krea 2 but left the matching check on the generate endpoint, so
    Generate failed with INVALID_REQUEST: Required fields are missing (duration_seconds).
    Duration is now a target capability everywhere it is checked: required only for the
    targets that declare one, and never read as if it always exists.
  • Image prompts no longer carry a Duration: line at all, instead of the literal
    None seconds that could be read as a constraint.

Requirements

  • ComfyUI
  • Python 3.10 or newer
  • At least one configured provider

Models, API keys, and llama.cpp binaries are not bundled.

Looking for the version that runs without ComfyUI? See the
Standalone guide.

Prompt Studio for ComfyUI v1.1.3

Choose a tag to compare

@tngklp tngklp released this 27 Sep 12:14

Prompt Studio for ComfyUI v1.1.3

ComfyUI extension release. Install through ComfyUI Manager or extract the ZIP into
custom_nodes.

Download

Download Prompt-Studio-ComfyUI-v1.1.3.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.

Install

  1. Extract the ZIP so that the folder ComfyUI-Prompt-Studio sits inside
    ComfyUI/custom_nodes.
  2. Restart ComfyUI.
  3. Open the floating Prompt Studio button, or use Extensions > Prompt Studio.
  4. Open Settings and choose a provider.

ComfyUI Manager is the simpler path and handles updates. Use the ZIP only for a manual
install.

What's new

  • Fixed refining a prompt for an image target. Refine failed for Anima, Qwen Image 2.1
    and Krea 2 with INVALID_DURATION: The selected mode does not accept a duration.
    Generation and refinement both validated a video-only duration for every target, so the
    three image targets - which have no duration at all - were rejected before the request
    reached their guide. Duration is now validated only for the targets that declare one.
  • The duration line is omitted from an image request instead of being written as
    None seconds, so the prompt model no longer reads an absent field as a constraint.
  • An image target that is sent a duration anyway is still rejected with a clear error,
    rather than silently ignoring a field that does not apply to it.

Requirements

  • ComfyUI
  • Python 3.10 or newer
  • At least one configured provider

Models, API keys, and llama.cpp binaries are not bundled.

Looking for the version that runs without ComfyUI? See the
Standalone guide.

Prompt Studio Standalone v1.2.7 (Windows)

Choose a tag to compare

@tngklp tngklp released this 27 Sep 22:43

Prompt Studio Standalone v1.2.7

Standalone Windows version. ComfyUI is not required.

Download

Download Prompt-Studio-Standalone-Windows-v1.2.7.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.
Do not install this package into ComfyUI custom_nodes.

Updating

Run update.bat in the install folder:

update.bat

It downloads the latest release, replaces the application files, and keeps your data\,
models\, .venv\ and any llama-server or CUDA files you added. Your previous version
is backed up to a temp folder, so a bad update can be undone. Run start.bat afterwards.

What's new

  • Anima no longer assumes every character is a girl. Selecting one female and one
    male character wrote 2girls, because each character's gender was discarded before the
    prompt was built. A one-girl-one-boy selection now writes 1girl, 1boy.
  • The fix applies without deleting anything. The character catalogue records a
    format version, so an out-of-date cache is refreshed on the next launch instead of
    waiting up to 30 days for the normal refresh.
  • Requires Prompt Studio core 1.1.7.

ComfyUI-only controls, including Auto VRAM management and Add to workflow, are not
shown in Standalone.

Features

  • Ollama
  • API providers
  • External llama.cpp
  • Existing local GGUF models through a user-selected llama-server.exe
  • Vision GGUF projectors with metadata-based matching
  • Combined and removable model locations
  • Local GGUF Context, KV cache, Generation budget, and supported reasoning effort controls
  • Video Creative Briefs up to 8,000 characters

External llama.cpp keeps its own reasoning and chat-template settings. Prompt Studio separates returned reasoning from the final prompt without overriding the server.

No ComfyUI installation is required. Models, API keys, llama.cpp binaries, and CUDA
libraries are not bundled.

Requirements

  • Windows 10 or 11, x64
  • Python 3.10 or newer, or uv
  • At least one configured provider

Based on Prompt Studio extension 0.4.6.

Prompt Studio Standalone v1.2.6 (Windows)

Choose a tag to compare

@tngklp tngklp released this 27 Sep 22:30

Prompt Studio Standalone v1.2.6

Standalone Windows version. ComfyUI is not required.

Download

Download Prompt-Studio-Standalone-Windows-v1.2.6.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.
Do not install this package into ComfyUI custom_nodes.

Updating

Run update.bat in the install folder:

update.bat

It downloads the latest release, replaces the application files, and keeps your data\,
models\, .venv\ and any llama-server or CUDA files you added. Your previous version
is backed up to a temp folder, so a bad update can be undone. Run start.bat afterwards.

What's new

  • Anima's quality tags are now masterpiece, best quality, highres, score_9.
    Previously masterpiece, best quality, score_7. This applies to the system prompt and
    the starter prompt in the Anima workspace.
  • Anima no longer assumes every character is a girl. Selecting one female and one
    male character wrote 2girls. The subject count tag is now stated explicitly and
    derived from each character's own gender, so a mixed pair is 1girl, 1boy.
  • Requires Prompt Studio core 1.1.6.

ComfyUI-only controls, including Auto VRAM management and Add to workflow, are not
shown in Standalone.

Features

  • Ollama
  • API providers
  • External llama.cpp
  • Existing local GGUF models through a user-selected llama-server.exe
  • Vision GGUF projectors with metadata-based matching
  • Combined and removable model locations
  • Local GGUF Context, KV cache, Generation budget, and supported reasoning effort controls
  • Video Creative Briefs up to 8,000 characters

External llama.cpp keeps its own reasoning and chat-template settings. Prompt Studio separates returned reasoning from the final prompt without overriding the server.

No ComfyUI installation is required. Models, API keys, llama.cpp binaries, and CUDA
libraries are not bundled.

Requirements

  • Windows 10 or 11, x64
  • Python 3.10 or newer, or uv
  • At least one configured provider

Based on Prompt Studio extension 0.4.6.

Prompt Studio Standalone v1.2.5 (Windows)

Choose a tag to compare

@tngklp tngklp released this 27 Sep 14:20

Prompt Studio Standalone v1.2.5

Standalone Windows version. ComfyUI is not required.

Download

Download Prompt-Studio-Standalone-Windows-v1.2.5.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.
Do not install this package into ComfyUI custom_nodes.

Updating

Run update.bat in the install folder:

update.bat

It downloads the latest release, replaces the application files, and keeps your data\,
models\, .venv\ and any llama-server or CUDA files you added. Your previous version
is backed up to a temp folder, so a bad update can be undone. Run start.bat afterwards.

This is the first release you can update with update.bat alone. Updating to 1.2.5
from 1.2.4 still needs that one manual ZIP extract, because the copy of update.bat
being replaced is the one doing the replacing.

What's new

  • Fixed "Generate prompt" for every image target. Qwen Image 2.1 (T2I and Edit),
    Krea 2 and Anima returned INVALID_DURATION: The selected mode does not accept a duration. Video and music prompts were never affected.
  • Image targets now ignore the duration value the studio sends with every request, instead
    of rejecting the request for carrying a field it cannot use.
  • Requires Prompt Studio core 1.1.5.

ComfyUI-only controls, including Auto VRAM management and Add to workflow, are not
shown in Standalone.

Features

  • Ollama
  • API providers
  • External llama.cpp
  • Existing local GGUF models through a user-selected llama-server.exe
  • Vision GGUF projectors with metadata-based matching
  • Combined and removable model locations
  • Local GGUF Context, KV cache, Generation budget, and supported reasoning effort controls
  • Video Creative Briefs up to 8,000 characters

External llama.cpp keeps its own reasoning and chat-template settings. Prompt Studio separates returned reasoning from the final prompt without overriding the server.

No ComfyUI installation is required. Models, API keys, llama.cpp binaries, and CUDA
libraries are not bundled.

Requirements

  • Windows 10 or 11, x64
  • Python 3.10 or newer, or uv
  • At least one configured provider

Based on Prompt Studio extension 0.4.6.

Prompt Studio Standalone v1.2.4 (Windows)

Choose a tag to compare

@tngklp tngklp released this 27 Sep 14:07

Prompt Studio Standalone v1.2.4

Standalone Windows version. ComfyUI is not required.

Download

Download Prompt-Studio-Standalone-Windows-v1.2.4.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.
Do not install this package into ComfyUI custom_nodes.

Updating

Run update.bat in the install folder. It downloads the latest release, replaces the
application files, and keeps your data\, models\, .venv\ and any llama-server or
CUDA files you added. Run start.bat when it finishes. Your previous version is backed up
to a temp folder, so a bad update can be undone.

From 1.2.5 onward update.bat is included in the package. This release is the last one
that has to be updated by extracting the ZIP by hand
, because the running copy of
update.bat is the one the updater replaces.

What's new

  • Fixed generating a prompt for an image target. Generate failed for Anima, Qwen Image
    2.1 and Krea 2 with INVALID_REQUEST: Required fields are missing (duration_seconds).
    The 1.2.3 release fixed Refine for those targets but left the matching check on the
    generate path. Duration is now required only for the targets that declare one.
  • Image prompts no longer carry a Duration: line, instead of the literal None seconds.
  • update.bat for one-step updates. Run it in the install folder to move to the latest
    release without re-downloading or re-extracting anything by hand.
  • Requires Prompt Studio core 1.1.4.

ComfyUI-only controls, including Auto VRAM management and Add to workflow, are not
shown in Standalone.

Features

  • Ollama
  • API providers
  • External llama.cpp
  • Existing local GGUF models through a user-selected llama-server.exe
  • Vision GGUF projectors with metadata-based matching
  • Combined and removable model locations
  • Local GGUF Context, KV cache, Generation budget, and supported reasoning effort controls
  • Video Creative Briefs up to 8,000 characters

External llama.cpp keeps its own reasoning and chat-template settings. Prompt Studio separates returned reasoning from the final prompt without overriding the server.

No ComfyUI installation is required. Models, API keys, llama.cpp binaries, and CUDA
libraries are not bundled.

Requirements

  • Windows 10 or 11, x64
  • Python 3.10 or newer, or uv
  • At least one configured provider

Based on Prompt Studio extension 0.4.6.

Prompt Studio Standalone v1.2.3 (Windows)

Choose a tag to compare

@tngklp tngklp released this 27 Sep 12:14

Prompt Studio Standalone v1.2.3

Standalone Windows version. ComfyUI is not required.

Download

Download Prompt-Studio-Standalone-Windows-v1.2.3.zip from the release assets below.

Do not download Source code (zip) or Source code (tar.gz) for normal use.
Do not install this package into ComfyUI custom_nodes.

What's new

  • Fixed refining a prompt for an image target. Refine failed for Anima, Qwen Image 2.1
    and Krea 2 with INVALID_DURATION: The selected mode does not accept a duration.
    Duration was validated for every target, so the three image targets - which have no
    duration at all - were rejected before the request reached their guide.
  • The duration line is omitted from an image request instead of being written as
    None seconds, so the prompt model no longer reads an absent field as a constraint.
  • Requires Prompt Studio core 1.1.3.

ComfyUI-only controls, including Auto VRAM management and Add to workflow, are not
shown in Standalone.

Features

  • Ollama
  • API providers
  • External llama.cpp
  • Existing local GGUF models through a user-selected llama-server.exe
  • Vision GGUF projectors with metadata-based matching
  • Combined and removable model locations
  • Local GGUF Context, KV cache, Generation budget, and supported reasoning effort controls
  • Video Creative Briefs up to 8,000 characters

External llama.cpp keeps its own reasoning and chat-template settings. Prompt Studio separates returned reasoning from the final prompt without overriding the server.

No ComfyUI installation is required. Models, API keys, llama.cpp binaries, and CUDA
libraries are not bundled.

Requirements

  • Windows 10 or 11, x64
  • Python 3.10 or newer, or uv
  • At least one configured provider

Based on Prompt Studio extension 0.4.6.