One API for hundreds of AI models with unified billing and refundable balance
Stage
Established
Startup
Founded
Not given
2026
Based in
US
Not given
Pricing model
Not given
Paid
Pricing
Not given
Usage-based prepaid credits with transparent per-token pricing. Examples: GPT-6.1 Terra from $0.15 input / $0.75 output per 1M tokens; Claude Haiku 5.5 from $0.03/M input / $0.12/M output; Claude Opus 5.5 from $1.12/M input / $5.6/M output. First top-up receives 10% bonus credits (up to $200, minimu
Key features
GroqMetal: dedicated bare-metal infrastructure tuned for speed and reliability
GroqCore: production-ready inference stack with no infrastructure expertise required
GroqAssured: enterprise-grade governance, auditability, and control
256 LPUs per rack with 40 PB/s SRAM bandwidth
1,000 tokens per second per user throughput
128 GB on-chip SRAM per rack
315 PFLOPS of FP8 inference compute
13 globally distributed data centers across four continents
One API for hundreds of models across OpenAI, Anthropic, Google, DeepSeek, MiniMax and more
OpenAI, Anthropic, and Vertex protocol compatibility
Multi-layer redundancy with automatic failover
Usage-based billing with no monthly minimum
Only successful requests are billed, failures are free
Refundable balance at any time
Visible rate limits per account
Support for text, image and video workloads
What makes it different
Fast inference at scale without sacrificing affordability
LPU pioneered technology integrated with NVIDIA GPUs
Fully integrated platform combining infrastructure, inference, and control
Reliability-first architecture with multi-layer redundancy across network, gateway and model layers
Transparent, refundable balance instead of locked-in subscriptions
Unified billing across all model families with one invoice
Health-aware routing across vendors with automatic failover
RouteMux