Professional multi-modal workspace for Puter.js AI.
Ready
AI ConversationReady for a new conversation
✦
How can I help?Ask anything, write code, analyze ideas, or work through a problem.
Optional image or video · maximum 100 MB per request. You can send normal text without an attachment.
PROMPT0 tokens
CONVERSATION0 tokens
MAX RESPONSE2,000
CONTEXT LEFT—
⚙ Generation controlsFine-tune the next response
Loading every available provider…Checking capabilities…
Sampling control for models that support temperature.
Used by supported reasoning models.
✦
Image conversationYour prompts and generated images
▧Create your first imageThe generated image will appear here.
＋Image referenceOptional URLs, separated by commas
◉
Vision conversationImage understanding and OCR results
◉Analyze an imageYour image and AI analysis will appear here.
↑Upload imageChoose a local image up to 10 MB
Use an image URL or a local image. OCR extracts text; Analyze understands the full image.
♪
Voice conversationText prompts and playable speech
♪Generate your first voiceThe audio player will appear here when ready.
▶
Video conversationPrompts and generated video clips
▶Generate your first videoThe video player will appear here when ready.
Video availability depends on your provider. Backend errors are shown directly in this conversation.
▤
File conversationDocuments and saved AI analysis
▤Analyze a documentThe document request and answer will appear here.
↑Upload documentSelect one local file for this request
The temporary cloud copy is removed after the request finishes. Support depends on the selected model.
Click a button to query Puter's currently exposed model/provider catalog.
SIGN-INChecking…
USER—
MONTHLY USAGE—
SDKPuter.js v2
Authentication is handled by Puter. Do not paste OpenAI API keys into this page.
TOKEN INTELLIGENCE
Usage center
Understand how your prompt, conversation history, and requested response share the model context window.
CURRENT PROMPT0estimated tokens
CHAT HISTORY0estimated tokens
RESPONSE BUDGET2,000maximum requested
MODEL CONTEXT—catalog metadata
Puter monthly usage
Sign in to load the usage information supplied by your Puter account.
What is a token?
A token is a small piece of text. In English, one token is roughly four characters or about three-quarters of a word. Other languages and code can use tokens differently.
What consumes tokens?
Your new message, recent conversation history, system instructions, attached content, tool results, and the generated answer all consume context.
Why context matters
The prompt and answer must fit inside the model’s context window. Longer chats leave less room for the next response.
Estimate versus billing
Counts shown here are browser estimates. The provider’s account usage is authoritative and may use credits, tokens, or another unit depending on the model.
PERSONALIZATION
Workspace settings
Control the look, chat behavior, saved account state, and local privacy preferences.
Save account session
Keep the Puter sign-in session available through Puter's own authentication system. This app never stores your password.
Save conversations locally
Store chat history in this browser using localStorage.
Auto-focus message box
Focus the composer whenever you start a new chat.
Account authentication remains controlled by Puter. “Save account” means this app remembers your preference; it does not save credentials.
⚙ Function calling
Run a small calculator tool through the model. The browser executes the requested function and sends its result back for a final response.
Delete conversation?
This permanently removes the selected chat from this browser.