Show HN: NvChat – A single-exe Windows client for Nvidia's free LLM API

akon471 pts0 comments

GitHub - akon47/NvChat · GitHub

/" data-turbo-transient="true" />

Skip to content

Search or jump to...

Search code, repositories, users, issues, pull requests...

-->

Search

Clear

Search syntax tips

Provide feedback

--><br>We read every piece of feedback, and take your input very seriously.

Include my email address so I can be contacted

Cancel

Submit feedback

Saved searches

Use saved searches to filter your results more quickly

-->

Name

Query

To see all available qualifiers, see our documentation.

Cancel

Create saved search

Sign in

/;ref_cta:Sign up;ref_loc:header logged out"}"<br>Sign up

Appearance settings

Resetting focus

You signed in with another tab or window. Reload to refresh your session.<br>You signed out in another tab or window. Reload to refresh your session.<br>You switched accounts on another tab or window. Reload to refresh your session.

Dismiss alert

{{ message }}

akon47

NvChat

Public

Notifications<br>You must be signed in to change notification settings

Fork

Star

main

BranchesTags

Go to file

CodeOpen more actions menu

Folders and files<br>NameNameLast commit message<br>Last commit date<br>Latest commit

History<br>24 Commits<br>24 Commits

.github/workflows

.github/workflows

NvChat

NvChat

docs

docs

.gitignore

.gitignore

LICENSE

LICENSE

NvChat.sln

NvChat.sln

README.ko.md

README.ko.md

README.md

README.md

View all files

Repository files navigation

NvChat

A Windows desktop app to chat with the free LLMs on build.nvidia.com

Streaming chat · saved conversations · live markdown/code rendering · reasoning display · image attachments · in-app auto-update — ships as a single .exe

English · 한국어

Markdown renders live as the answer streams in.

✨ Features

Model picker — loads the live model list via /v1/models (falls back to a curated list on failure)

Streaming chat — real-time token streaming with an instant Stop button

Live markdown while streaming — formatting renders as the answer arrives (like Claude / ChatGPT), not only when it finishes

Reasoning display — collapsible thinking for models that emit reasoning_content / (deepseek-r1, etc.)

Image attachments (vision) — attach images to ask vision models (llama-3.2-vision, etc.); large images are auto-downscaled

In-app auto-update — checks GitHub Releases, downloads the new .exe, verifies its SHA-256, and swaps itself in place (single-file friendly, no installer)

Per-conversation usage — tracks requests and prompt/completion tokens per model , so you can see what each model spent in a conversation (even across mid-chat model switches)

Desktop launcher — a global hotkey (default Ctrl+Shift+Space) pops a quick-chat mini window from anywhere + system tray (close minimizes to tray)

Personalization — custom instructions (about you / response style) applied to every chat + reusable prompt presets

Searchable model picker — filter 100+ models by typing

Message actions — regenerate a response, edit & resend a user message, delete or copy individual messages

Selectable markdown — headings/lists/nested & task lists/quotes/tables /links, syntax-highlighted code blocks with a copy button; assistant text is drag-selectable

Conversation management — sidebar search , date groups (Today/Yesterday/…), pin , rename , autosave, model-generated auto titles

Per-conversation settings — system prompt, Temperature/Top P/Max Tokens/Penalties

Export — copy the whole conversation or save as Markdown

Quality of life — scroll-to-bottom button, remembers window size/position, collapsible sidebar, keyboard shortcuts

Safe storage — API key encrypted with Windows DPAPI, atomic file writes + automatic backup of corrupt files

Multi-language — English & Korean UI, switchable in Settings (defaults to your Windows language on first run)

UI — custom borderless dark theme with an NVIDIA-green accent; ChatGPT/Claude-style layout (bubble only on your messages)

📦 Download

Grab the single NvChat.exe from Releases and run it.<br>It's a self-contained single file that runs even without the .NET runtime installed. (Windows x64)

First run: the app isn't code-signed, so Windows SmartScreen may show a blue "Windows protected your PC" prompt. Click More info → Run anyway .

Once you're on a build with auto-update, new versions are offered inside the app — no need to re-download manually.

🔑 Get an API key

Sign in to build.nvidia.com

On any model page, click Get API Key to create an nvapi-... key

Paste it into the Settings window on first launch (or the ⚙ button), then Test connection → Save

The key is stored only on this PC + this Windows account , encrypted with DPAPI (%APPDATA%\NvChat\settings.json). It cannot be decrypted on another PC/account.

⌨️ Shortcuts

Key<br>Action

Enter<br>Send (switchable to Ctrl+Enter in Settings)

Shift+Enter<br>New line

Ctrl+N<br>New conversation

Ctrl+Shift+Space<br>Quick chat from anywhere (global · configurable)

🛠️ Build from source

# Run for development<br>dotnet run --project NvChat/NvChat.csproj

# Build the single-file...

nvchat windows model settings chat single

Related Articles