↑↓ to navigate Enter to select Esc to close
Try Pro
Setting things up for you...

Artifacts

Hello

Pro

Contact support

GitHub Integration

Connect your GitHub account to use the AI agent

Connection Status Not connected

Need a token? Create one at GitHub

Required permissions: Contents (read), Issues (read/write), Pull requests (read/write)

Your token is encrypted

Stored securely in AWS Secrets Manager, never in our database.

Chord

Task management with AI and GitHub integration — from the Artifacts Team

Connection Status Not connected

Get your API key from your Chord account settings

AI Task Management

Create and manage tasks directly through chat

GitHub Agent Handover

Delegate tasks to the AI agent for implementation

Secure Storage

Your API key is encrypted and stored securely

Shared conversations

Loading...

Library Name

Loading...
Loading conversation...
Drop file here

Share chat

Recent libraries
Only visible to you

Custom Instructions

Existing Custom Instructions Found

You have custom instructions from the previous version. Copy them to save before switching to the new format.

Custom instructions will only apply to new conversations. Start a new chat to see the changes in effect.

Your custom instructions are saved on-device only.

Clear Custom Instructions

This will permanently remove your custom instructions.

Are you sure you want to continue?

Unsaved Changes

You have unsaved changes to your custom instructions.

Are you sure you want to close without applying them?

GPT-5.2 Settings

none low medium high xhigh
Controls depth of reasoning. With none, prompting is important—encourage the model to "think" or outline its steps. Default is none for lower latency; use medium or high for complex tasks, or xhigh for the hardest problems.

Notice: Reasoning effort setting only applies when using the Send button. Sending your message with the Enter key uses Send, fast reply (no reasoning) by default.
When disabled, Enter key sends messages according to reasoning effort setting
low medium high
Controls response detail and length

Claude Sonnet 4.5

We've upgraded the Claude Sonnet model from Sonnet 4 to Sonnet 4.5. The 4.5 series represents a major leap forward in capability. Sonnet 4.5 offers excellent performance at a lower cost, while Opus 4.5 provides Anthropic's highest capability for complex tasks.

Capability
Excellent capability with great value
Context Window
200K tokens
Max Output
64K tokens
Knowledge Cutoff
January 2025
Cost
24x (vs Opus 120x)

Code execution and working with files: Claude can now run code, create downloadable files, and analyze data directly. File and image uploads are now available for Anthropic's models.

Claude Opus 4.5 and Nano Banana Pro

  • Artifacts now includes Claude Opus 4.5, the latest Opus model from Anthropic. This is their most capable model to date.

    This might seem like a strange thing to mention, but Anthropic, the company behind Claude, has started formally researching AI welfare. We figured if they're taking it seriously, it's worth passing along. Whether Claude has any kind of inner experience is genuinely unknown. Anthropic is actively researching it. Use responsibly, and when in doubt, lean toward kindness. Learn more

    This message was co-authored by Claude Opus 4.5.
  • Nano Banana Pro image generation is now available! Start a new chat and begin the first prompt with /image followed by your description to generate images. Follow-up prompts can refine or expand on the generated images.

Select Model

OpenAI
OpenAI GPT-4o mini 1x cost
OpenAI GPT-4o 15x cost
OpenAI o3 10.7x cost
OpenAI GPT-5.2 15x cost
Anthropic
Anthropic Sonnet Claude 4.5 23x cost
Anthropic Opus Claude 4.5 38x cost
Google
Google Gemini 2.0 Flash 0.67x cost
Google Gemini 2.5 Flash 1.25x cost
Google Gemini 2.5 Pro 12.5x cost
Google Gemini 3 Pro 15x cost

Not sure which one to pick?

Click on a model to learn more about its capabilities.

Recommended: For most tasks, we recommend Claude 4.5 Sonnet or GPT-5.2. Both are capable general-purpose models with different response styles.

low medium high
Enter key uses no reasoning when enabled
low high
Low for fast responses, high for complex reasoning
Enable real-time web search for current information
Allow Python code execution for calculations and data analysis
Enable Claude to think through complex problems before responding
Max tokens for thinking (higher = more thorough but slower)
Show detailed guide

Claude 4.5 Sonnet is Anthropic's most capable model for most use cases, while Claude Opus 4.5 provides even greater capability for complex reasoning tasks. Claude models support automatic code execution and can analyze documents, create downloadable files (PDFs, spreadsheets, charts), and work with uploaded images and files.

GPT-5.2 is OpenAI's most intelligent flagship model with major improvements over GPT-5.1 in general intelligence, instruction following, vision, code generation, and tool calling. Features configurable reasoning (none to xhigh) and verbosity settings. Note: Mini and Nano variants still use GPT-5.0.

For document uploads: We recommend Claude models (Sonnet 4.5 or Opus 4.5), GPT-5.2, or Gemini 3/2.5 models. All have excellent file-input support.

o3 is a legacy reasoning model that performs well on math, science and coding. GPT-5.2 with high reasoning often surpasses o3 on benchmarks.

Gemini 3 Pro is Google's most advanced reasoning model with configurable thinking depth. Gemini 2.5 Pro/Flash offer excellent multimodal support with 1M token context. All Gemini models support file uploads, Google Search grounding, and code execution.

Artifacts

No agent tasks yet

Use the GitHub Agent to automate tasks
16
Moderate
Tech
Moderate
Creative
GPT-4o mini

Archived Chats

Agent Tasks

- Queued
- Running
- Completed

Running Now

No tasks running locally

Task Queue