Model lineup
Real models, honest names.
Every workspace choice is one slug that works in the chat, the API and MCP alike. We publish which model is behind it, what it costs per token, and which plans unlock it - compare freely.
Swift
fastFree planOpenAI gpt-oss-20b on Groq. Fast, low-cost chat for everyday questions and drafts.
Everyday questions, quick drafts, translation, summaries - fastest and cheapest.
- per 1k in
- 2 cr≈₹0.02
- per 1k out
- 8 cr≈₹0.08
- max reply
- 8,192tokens
A typical reply (~500 tokens in, ~300 out) costs about 3.4 credits ≈ ₹0.03.
Balanced
balancedRecommendedPaid plansOpenAI gpt-oss-120b on Groq. Stronger reasoning for writing, analysis and code.
Writing, analysis and code - the all-rounder for real work.
- per 1k in
- 4 cr≈₹0.04
- per 1k out
- 16 cr≈₹0.16
- max reply
- 8,192tokens
A typical reply (~500 tokens in, ~300 out) costs about 6.8 credits ≈ ₹0.07.
The same slugs work everywhere: pick them in the workspace, or send model: "fast" to the API. Rates are charged per token actually used - failed requests refund in full. API docs · plans & top-ups · what replies really cost