The Inference Directory
Explore directory

Sarvam AI

Verified

Sarvam AI runs an Indian token factory with multilingual LLMs, speech, OCR, and translation APIs, hosted in India.

Indian-owned India hosting Token factory
Visit website API docs

At a glance

Record
Legal entity
Axonwise Private Limited
Headquarters
Bengaluru, Karnataka, India
Hosting region
India (Bengaluru)
Group
Token factory
SLA
99.95%

Summary

Scope

Sarvam AI runs an Indian token factory with multilingual LLMs, speech, OCR, and translation APIs, hosted in India.

Prices exclude 18% GST. LLM inference runs on Yotta and NxtGen GPUs.

Source: www.sarvam.ai ↗ · Updated Nov 4, 2026

API access

OpenAI-compatible

api-subscription-key (or Bearer)

Billing

PAYG, Pro ₹9,999, Growth ₹49,999, Business ₹1,99,999, Enterprise custom. ₹100 free credits.

KYC

Business account

Compliance

ISO 27001SOC 2DPA

Key models

23 listed
Sarvam 105B 128K · per 1M tokens LLM ₹15 / ₹60 per 1M tokens
Sarvam 105B Conversations 32K · per 1M tokens LLM ₹15 / ₹60 per 1M tokens
GLM-5.3 1M · per 1M tokens LLM ₹126 / ₹390 per 1M tokens
DeepSeek V4 Pro 1M · per 1M tokens LLM ₹115 / ₹350 per 1M tokens
GLM-5.3 Flash 1M · per 1M tokens Multimodal ₹13.5 / ₹45 per 1M tokens
Gemma 4 31B 256K · per 1M tokens Multimodal ₹35 / ₹90 per 1M tokens
DeepSeek V4.1 Flash 1017 1M · per 1M tokens Multimodal ₹25 / ₹105 per 1M tokens
Kimi K3 1M · per 1M tokens Multimodal ₹270 / ₹1350 per 1M tokens
MiMo V2.6 Flash 1M · per 1M tokens Multimodal ₹12.5 / ₹25 per 1M tokens
MiMo V2.6 Pro 1M · per 1M tokens Multimodal ₹40 / ₹80 per 1M tokens
Sarvam Prism Context n/p · per 1M tokens Decision ₹4 / ₹0 per 1M tokens
Saaras v4 Context n/p · per minute STT ₹0.49 per minute
Saaras v3 Context n/p · per minute STT ₹0.49 per minute
Saarika v2.5 Context n/p · per minute STT ₹0.49 per minute
Diarization (batch) Context n/p · per minute STT ₹0.59 per minute
Bulbul v4 Flash Context n/p · per 1k chars TTS ₹1 per 1k chars
Bulbul v3 Context n/p · per 1k chars TTS ₹3 per 1k chars
Instant Voice Cloning Context n/p · per 1k chars TTS ₹2 per 1k chars
Sarvam Vision , digitization Context n/p · per page OCR ₹0.5 per page
Sarvam Vision , extraction Context n/p · per page OCR ₹1 per page
Sarvam Translate / Mayura Context n/p · per 1k chars Translation ₹2 per 1k chars
Transliteration Context n/p · per 1k chars Translation ₹2 per 1k chars
Language identification Context n/p · per 1k chars Utility ₹0.35 per 1k chars

Pricing

List rates · local currency · GST excluded
Model Type Input price Caching price Output price
Sarvam 105B128K · per 1M tokens LLM ₹15 ₹5 ₹60
Sarvam 105B Conversations32K · per 1M tokens LLM ₹15 ₹5 ₹60
GLM-5.31M · per 1M tokens LLM ₹126 ₹20 ₹390
DeepSeek V4 Pro1M · per 1M tokens LLM ₹115 ₹4 ₹350
GLM-5.3 Flash1M · per 1M tokens Multimodal ₹13.5 ₹2.5 ₹45
Gemma 4 31B256K · per 1M tokens Multimodal ₹35 ₹10 ₹90
DeepSeek V4.1 Flash 10171M · per 1M tokens Multimodal ₹25 ₹0.5 ₹105
Kimi K31M · per 1M tokens Multimodal ₹270 ₹27 ₹1350
MiMo V2.6 Flash1M · per 1M tokens Multimodal ₹12.5 ₹0.25 ₹25
MiMo V2.6 Pro1M · per 1M tokens Multimodal ₹40 ₹0.3 ₹80
Sarvam PrismContext n/p · per 1M tokens Decision ₹4 - ₹0
Saaras v4Context n/p · per minute STT ₹0.49 - -
Saaras v3Context n/p · per minute STT ₹0.49 - -
Saarika v2.5Context n/p · per minute STT ₹0.49 - -
Diarization (batch)Context n/p · per minute STT ₹0.59 - -
Bulbul v4 FlashContext n/p · per 1k chars TTS ₹1 - -
Bulbul v3Context n/p · per 1k chars TTS ₹3 - -
Instant Voice CloningContext n/p · per 1k chars TTS ₹2 - -
Sarvam Vision , digitizationContext n/p · per page OCR ₹0.5 - -
Sarvam Vision , extractionContext n/p · per page OCR ₹1 - -
Sarvam Translate / MayuraContext n/p · per 1k chars Translation ₹2 - -
TransliterationContext n/p · per 1k chars Translation ₹2 - -
Language identificationContext n/p · per 1k chars Utility ₹0.35 - -

List rates. Cached input shown where published; "-" means the provider does not publish a cache rate. Prices exclude 18% GST where applicable. PAYG, Pro ₹9,999, Growth ₹49,999, Business ₹1,99,999, Enterprise custom. ₹100 free credits.

Infrastructure

Source-backed
RegionIndia
CitiesBengaluru
SLA99.95%
API basehttps://api.sarvam.ai/v2
Authapi-subscription-key (or Bearer)
Rate limitPAYG 50 RPM; Pro 200 RPM; Growth 1,000 RPM.

Access

Billing & API
API basehttps://api.sarvam.ai/v2
Authapi-subscription-key (or Bearer)
Rate limitPAYG 50 RPM; Pro 200 RPM; Growth 1,000 RPM.
SLA99.95%

PAYG, Pro ₹9,999, Growth ₹49,999, Business ₹1,99,999, Enterprise custom. ₹100 free credits.

Compliance

Source-backed
DPAAvailable
ZDRon_request
Trainingno cross-customer training; consumer policy opt-in by default
RetentionZDR on request; default retention is set by the account owner; DPA available.
Certifications
ISO 27001:2022SOC 2 Type ISOC 2 Type II
Subprocessors

Microsoft Azure Central India, Yotta, NxtGen

Benchmarks

Coming soon

Coming soon.

Performance

Coming soon
Latency

Time to first token

Median and p95 response times for short and long prompts, measured from repeatable probes rather than vendor claims.

Coming soon
Throughput

Tokens per second

Streaming output rate and request throughput under steady load, with concurrency and context length recorded.

Coming soon
Efficiency

Cost per workload

Blended cost for a fixed workload, including cache reads and output tokens, so speed and price can be compared together.

Coming soon

Independent performance probes are being added. Vendor-reported numbers are not mixed into this section.