Model catalogue
Every model, direct. No route aliases.
You pick the exact open-weight model; we publish its card, guarantee the SLA and residency, and announce version changes 60 days ahead. Model origin is not hosting location: every model serves from Indian GPUs, and no request is ever routed to a China-hosted or offshore inference API.
4 models operational· all systems normal
ModelBest useContextPricing
Chat
DeepSeek V4 Flash (alias, now V4.1)
deepseek-v4-flash · Chat
Chat, coding, agents, extraction, JSON
256K
in₹20/MTok$0.21/MTok
cached₹1/MTok$0.011/MTok
out₹60/MTok$0.63/MTok
DeepSeek V4.1 Flash
deepseek-v4.1-flash · Chat
Chat, coding, agents, extraction, JSON
256K
in₹20/MTok$0.21/MTok
cached₹1/MTok$0.011/MTok
out₹60/MTok$0.63/MTok
GLM-5.3 Flash
glm-5.3-flash · Chat
Chat, coding, agents, extraction, JSON
256K
in₹6.73/MTok$0.071/MTok
cached₹1.35/MTok$0.014/MTok
out₹22.43/MTok$0.24/MTok
Decision (System One)
Perplexity Decider v1.1 (27B)
pplx-decider-v1.1 · Decision (System One)
Routing, triage, classification, guardrails
8K
in₹8/MTok$0.084/MTok
Coming soon:Image generationVideo understanding
cached = input tokens served from the prefix cache, billed at that model's cached rate · Rates per million tokens, flat 24/7, ex-GST · A published card for every modelUSD rates are converted at ₹94.82 = $1 (ECB reference rate, 2026-10-22) and plan prices are rounded. All charges are billed in INR.