Thread Deck is a canvas-based workspace that puts your AI chats, personas, briefs, and decisions on one visual canvas, so you stop switching between tabs and pasting the same context again. You save reusable blocks, run them through models like GPT-5, Claude, and Gemini, and each run records the model, token use, and cost. It suits founders, creators, agencies, ops leads, and growth teams that need a clear record of how each AI decision was made.

• Visual Canvas Workspace: See every chat, persona, brief, and decision in one view instead of scattered tabs. A node graph and grouping into grids or clusters keep large projects readable.
• Reusable Blocks: Save chats, personas, briefs, notes, and prompt recipes as blocks. Pin them to the canvas and reuse them in new runs without retyping context.
• Multi-Model Access: Paid plans include GPT-5, Claude, and Gemini, so you can work with the model that fits each task from one place.
• Run Tracking: Every run logs the model used, input and output tokens, and the blocks that shaped the output. You get a clear record of how each answer was produced.
• Line-by-Line Cost Visibility: Runs show cost data and token multipliers, so you can see what each model call actually used from your allowance.
• Shareable Canvases: Share canvases through secure links with role-based permissions. Paid plans add richer sharing and collaboration controls.
• Canvas Analytics: Pro and higher plans include analytics to help you spot usage and spending spikes across canvases.
• AI Image Generation: Creator and Pro plans include access to AI image generation models alongside the text models.

• Less Context Repetition: Reusable blocks for personas, briefs, and prompts mean you set up context once and apply it to many runs.
• Clear Audit Trail: Each run records its model, tokens, and source blocks, which makes it easy to explain how an AI output was produced.
• Multiple AI Models in One Place: Paid plans bundle GPT-5, Claude, and Gemini, so you don't need separate subscriptions just to compare results.
• Transparent Spend: Per-run cost data and token multipliers show where your credits go before they run out.
• Free Plan to Start: You can try the canvas with 3 canvases and GPT-4o mini at no cost before choosing a paid tier.

• Very Limited Free Tier: The Free plan offers 1K blended tokens per month, GPT-4o mini only, and 3 canvases, so it works for a quick test rather than ongoing use.
• Token-Based Billing Takes Planning: Usage is measured in blended tokens with model-specific multipliers (for example, GPT-5 uses 1.5×), so your monthly allowance can shrink faster than expected on premium models. Overages cost $13 or $9 per 10K tokens depending on the plan.
• Team Pricing Is Not Public: Team and Enterprise plans have no published prices on the pricing page, so larger organizations need to talk to sales to estimate cost.
• View-Only Sharing on Free: Free workspaces get view-only share links, so real collaboration requires a paid plan.

• Agency Client Work: Keep each client's brand persona, briefs, and prompt recipes as blocks on one canvas. Your team reuses the same context across campaigns and can show clients how each deliverable was produced.
• Comparing Model Outputs: Run the same brief through GPT-5, Claude, and Gemini and keep the results next to each other. You pick the strongest output and see what each run cost.
• Documenting AI Decisions for Ops Teams: Ops leads and advisory teams can trace every decision back to the run, model, and inputs behind it. This gives you a clear record when stakeholders ask how a recommendation was made.
• Founder Research and Planning: Map out product ideas, market notes, and AI-generated analysis on one canvas instead of across dozens of chat tabs. Related threads stay linked, so you can pick up where you left off.
• Growth Team Content Pipelines: Store reusable prompt recipes for ads, emails, and posts, then run them as needed. Canvas analytics on Pro help you watch usage as volume grows.
• Sharing Work with Stakeholders: Send a secure canvas link with role-based permissions so teammates or clients can review the full thread without needing access to your AI accounts.