Playground
Model playground
Experiment with prompts, models and parameters.
Prompt
Test prompts against any model before shipping to production.
Press ⌘ + Enter to run
Response
velora-pro
142 tokens1.84s
Here's a concise launch plan for your AI writing assistant:
1. Positioning — lead with the job-to-be-done ("first drafts in your brand voice"), not the model.
2. Onboarding — ask for 3 sample documents, then generate a voice profile the user can edit.
3. Pricing — usage-based credits with a generous free tier; show remaining credits in-app.
4. Guardrails — log every request, rate-limit per API key and surface errors in a dashboard.
Want me to turn this into a 4-week roadmap?
Settings
Saved per workspace.
Most capable · 200k context
$3 / 1M in · $15 / 1M out
0.7
Higher values make output more random.
1024
Upper bound on generated tokens.
1
Nucleus sampling threshold.