FriendliAI
AutomatedAI inference platform offering model APIs, dedicated endpoints, and containers for fast and affordable LLM inference.
FriendliAI has 7 plans: paid plans from $0.1 and custom pricing for larger teams. We have not recorded a price change.
FriendliAI plans and prices
Model APIs (Pay-per-token)
usage based$0.1
From $0.10/1M tokens (Llama-3.1-8B-Instruct)
- Pay per token
- Frontier model inference
- GLM-5.1: $1.4 input/$4.4 output
- Llama-3.3-70B: $0.6
- DeepSeek-V3.2: $0.5 input/$1.5 output
- Whisper-large-v3: $0.0015/audio minute
Dedicated Endpoints A100
usage based$2.90
$2.9/hour (A100 80GB)
- Pay per second
- No start-up charges
- On-demand deployment
Dedicated Endpoints H100
usage based$3.90
$3.9/hour (H100 80GB)
- Pay per second
Dedicated Endpoints H200
usage based$4.50
$4.5/hour (H200 141GB)
- Pay per second
Dedicated Endpoints B200
usage based$8.90
$8.9/hour (B200 180GB)
- Pay per second
Container
Custom
- Run inference with full control
- Inference in your environment
Enterprise
Custom
- Custom Model APIs rate limits
- Priority GPU access
- Reserved GPU capacity
- VPC deployments
- On-prem deployment options
- Dedicated support
- Named Customer Success
Price history
Last 0 recorded changes
No price changes recorded yet.
Other AI & ML products
No confirmed alternatives yet.