The chat is the main place you'll work in Vamos. It's where you ask questions, paste links, upload files, and turn research into finished content. Here's every feature and when to use it.
Chat input anatomy
From left to right:
📎 File upload — attach PDFs, docs, spreadsheets, images, audio, video.
🔗 Add link — paste a YouTube, Instagram, TikTok, LinkedIn, Facebook, X, or website URL.
Text field — type your prompt.
🤖 Model selector — defaults to AUTO. Change it to force a specific model if you want.
➤ Send — run the prompt.
You don't have a Brand Profile or Brand Voice toggle anymore — Vamos loads your Brand Profile and Brand Voice automatically based on the task. Writing tasks always use your voice; research tasks don't need to.
Attaching files
Click the 📎 icon or drag a file in.
Supported formats:
Documents: PDF, TXT, DOCX, CSV, Excel
Images: PNG, JPG, JPEG, SVG
Videos: MP4 and most video formats
Audio: MP3, WAV, M4A and most audio formats
Limit: 10 MB per file, one at a time (multi-file upload coming soon).
What Vamos does with them:
PDFs / docs: extracts text, answers questions about content.
Images: describes what's in the image, extracts any text (OCR).
Videos: transcribes audio, analyzes visuals for shorter clips.
Audio: transcribes speech.
Spreadsheets: reads data, summarizes trends, can create outputs from rows.
Common uses:
Upload your own script → "Rewrite this to be punchier and cut 20 seconds."
Upload a competitor's PDF → "What makes this sales page effective?"
Upload your logo → "Describe my brand's visual style."
Adding links
Two ways: paste the URL directly into the text field, or click 🔗 and paste into the popup.
What works:
YouTube — individual videos (any length) or channel URLs (10–20 videos pulled).
Instagram — single post/Reel URLs or profile URLs. Reels get full transcription + visual analysis; image posts get caption + metrics.
TikTok — videos or profiles.
LinkedIn — posts or profiles.
Facebook — posts (including Ad Library links), pages.
X / Twitter — posts and profiles.
Any website — full page content extraction.
Private / region-restricted content won't load.
Full list: Supported social media links.
Model selector
Leave it on AUTO for 95% of tasks. Vamos picks Claude for writing, cheaper models for research and simple questions.
If you want to override — maybe force Claude Opus 4.6 for flagship writing, or Gemini 3.1 Pro for a massive context window — click the selector and pick manually. Each model shows its speed, quality, credit cost, and context window so you can pick confidently.
Full guide: AI Models: How to use them.
Guided forms for missing info
If your prompt is missing important details, Vamos will show a simple form asking for what it needs — platform, video length, audience, tone, CTA, etc. Fill it out and hit submit. Faster and clearer than a back-and-forth conversation.
Pre-filled next prompts
After Vamos finishes a response, it often suggests the next natural step as a pre-filled prompt. Just click send to keep moving — no typing needed.
Example: you ask for research. Vamos finishes with a pre-filled "Now write me 5 scripts based on these patterns" so you can continue the workflow in one click.
Editing messages + regenerating
Any message you've already sent can be edited. Click it, change the text, hit send — Vamos runs the updated prompt from that point forward in the conversation.
This is better than typing a "oh actually…" correction, because it replaces the old branch cleanly instead of adding confusion to the thread.
Stopping a generation
Click Stop while Vamos is generating. The response cancels instantly. No credits are consumed for the cancelled portion (you still pay for the partial work already done).
What Vamos can do in chat
You never have to tell it which tool to use — it figures it out from your prompt.
Search viral videos across TikTok, Instagram, YouTube
Analyze any public profile or post
Watch and transcribe videos
Pull content from any URL
Pull and analyze comments on social posts
Find similar accounts to any profile
Access proven viral hook frameworks
Search the web for current info
When Vamos researches, you see everything visually — thumbnails, cards, data — not just a text summary. Browse the source content before the final output.
See Research: In Chat for deeper examples.
When Canvas opens vs. when it doesn't
Canvas opens automatically when you ask for creative content: scripts, social posts, emails, ad copy, creative writing.
Canvas does NOT open for:
Research or analysis outputs
Short replies under 150 characters (shown as lists instead)
Lists of recommendations
Long-form articles formatted in markdown
Canvas details: Canvas: Edit your content.
Tips
Be specific, not long. "Write 3 TikTok scripts on X in my voice, 30 seconds each, hook = contrarian statement" beats five paragraphs of setup.
Chain research → writing in one prompt. Vamos will use a cheap model for research, then switch to Claude for writing inside the same response. Saves credits.
Correct in-line. If output's wrong, say "I hate this style — never use it again, save to memory."
New topic = new chat. Long chats accumulate context that can distract later tasks. Start fresh when you switch topics.
Questions? Chat widget or contact@getvamos.ai.