Starter
Pay as you go
Bonus creditsNo bonus credits
Rate limit60 requests/min
SupportCommunity support
Ideal forPrototyping & testing
Pricing
Simple plans for Serious model work.
Start with credits, move into a pilot, and scale into private deployment when your team is ready. No noisy plan maze.
Pro
INR 10,000
INR 1,000 bonus credits
Total credits11,000 credits
Rate limit200 requests/min
SupportEmail support
Ideal forStartups & POCs
Business
Most popularINR 50,000
INR 7,500 bonus credits
Total credits57,500 credits
Rate limit1,000 requests/min
SupportSolutions engineer
Ideal forProduction workloads
API Pricing
Usage is credit based across APIs. These are simple reference rates for planning pilots and production budgets.
API serviceDescriptionPriceUnit
Minnow-Math-1.5B
API service
Reasoning and math workloads with compact model latency.
INR 5 / 1M tokens
input or output
Reasoning-L1-10B
API service
Longer reasoning, structured Q&A, and agentic workflows.
INR 10 / 1M tokens
input or output
Minnow-OCR-1B
API service
Image analysis and document digitization for enterprise data.
INR 1
per page
Sonar-A1-STT
API service
Audio transcription with rounded per-second billing.
INR 30
per hour
Sonar-A1-TTS
API service
Voice generation charged by generated characters.
INR 15
per 10K chars
Minnow-Em1-0.6B
API service
Retrieval and semantic search vectors for private RAG.
INR 1
per 1M tokens
Minnow-VL-12B
API service
Vision-language understanding for multimodal workflows.
INR 15
per 1M tokens
FAQ
Questions, Answered
The short version of how plans, credits, and private deployment usually work.
Starter is the lightest path for prototypes. Move to Pro when you need predictable credits, higher limits, and support during evaluation.
Yes. Credits are universal across text, vision, speech, and embedding APIs unless your commercial agreement says otherwise.
Yes. Business and custom plans can include private cloud, VPC, on-prem, or air-gapped deployment planning.
We will move you to a custom plan with dedicated rate limits, deployment support, and model tuning scope.
Yes. We can structure pilots around data residency, audit needs, evaluation benchmarks, and internal approval cycles.
// Model access
Build with KiteFishAI models.
OCR, embeddings, reasoning, speech, vision-language, and math models are ready for evaluation and domain fine-tuning.
Industry-tuned
BFSI, healthcare & pharma
Base models
Available on Hugging Face
Pilot access
Co-develop domain fine-tunes