Model Catalogue
OpenAI-compatible raw model access, metered per token in rupees. Placeholder slots below — pricing and lanes are live once an engine is bound.
| Model | Tier | Modality | ₹/1M in | ₹/1M out | Context | Lane | Status |
|---|---|---|---|---|---|---|---|
| Edge 4B (slot) | edge | chat | 3 | 12 | 32K | S | waitlist |
| Small 8B (slot) | small | chat | 5 | 18 | 32K | S | waitlist |
| Mid 32B (slot) | mid | chat | 12 | 40 | 128K | R | waitlist |
| Large MoE (slot) | large-moe | chat | 20 | 60 | 256K | R | waitlist |
| Frontier Resident (slot) | frontier-resident | chat | 40 | 120 | 200K | G | waitlist |
| Code Mid (slot) | mid | chat | 12 | 40 | 64K | R | waitlist |
| apinfy Embed (BGE-M3) | small | embed | 2 | — | 8K | G | live |
| ASR Indic (slot) | small | asr | 6 | — | 1K | S | waitlist |
| TTS Indic (slot) | small | tts | 10 | — | 1K | S | waitlist |
| Image (slot) | mid | image | 0 | — | 1K | R | waitlist |
| Qwen 3.5 397B | large-moe | chat | 60 | 180 | 256K | G | live |
| Kimi K2.7 Code | large-moe | chat | 91 | 384 | 256K | G | live |
| DeepSeek V4 Flash | mid | chat | 42 | 127 | 128K | G | live |
| Qwen 3.8 Max | frontier-resident | chat | 192 | 576 | 256K | G | live |
| Qwen 3.7 Plus | large-moe | chat | 45 | 135 | 256K | G | live |
| Qwen 3.8 27B (India) | mid | chat | 30 | 90 | 16K | R | live |
| Ornith 1.0 35B (India) | large-moe | chat | 40 | 120 | 32K | R | waitlist |