AI Notebooks & Quantum Notebooks
CPU Units
For your testing, pre-production or basic machine learning phases, benefit from guaranteed CPU resources. You can deploy up to 12 resource units per notebook. Storage via Object Storage is unlimited.
Node | Memory | vCore | Ephemeral local storage | Public network | Price |
ai1-1-CPU | 4 GB | 1 | 40 GB | 500 Mbps | $0.042 |
GPUs
Implement GPU resources for all your compute-intensive AI training and deep learning needs.
Up to 4 resource units per notebook or workout can be deployed. Storage via Object Storage is unlimited.
Node | Memory | vCore | GPU | Ephemeral local storage | Public network | Price |
A100-1-GPU | 160 GB | 13 | NVIDIA Ampere A100 80GB 19.5 TFLOPS | 80 GB | 1.5 Gbps | $4.2 |
ai1-1-GPU | 40 GB | 13 | NVIDIA Tesla V100S 32GB 16.4 TFLOPS | 750 GB | 1.5 Gbps | $2.702 |
ai1-the-1-GPU | 40 GB | 13 | NVIDIA Tesla V100S 32GB 16.4 TFLOPS | 500 GB | 1.5 Gbps | $1.26 |
H100-1-GPU | 350 GB | 28 | NVIDIA Hopper H100 80GB 51 TFLOPS | 3072 GB | 5 Gbps | $4.34 |
L4-1-GPU | 80 GB | 20 | NVIDIA Ada Lovelace L4 24GB 30.3 TFLOPS | 500 GB | 5 Gbps | $1.162 |
L40S-1-GPU | 80 GB | 13 | NVIDIA Ada Lovelace L40S 48 GB 91.6 TFLOPS | 1536 GB | 5 Gbps | $2.17 |
The price is mentioned by the hour to make it easier to read the prices. The granularity of invoicing remains by the minute.
AI Training
CPU Units
Take advantage of guaranteed CPU resources, adapted to your test, pre-production or basic machine learning phases.
Deploy up to 12 resource units per notebook or workout. Storage via Object Storage is unlimited.
Node | Memory | vCore | Ephemeral local storage | Public network | Price |
ai1-1-CPU | 4 GB | 1 | 40 GB | 500 Mbps | $0.042 |
For reasons of simplification, the rate displayed is per hour. However, the granularity of invoicing is indeed minute.
GPUs
Deploy GPU resources for all your compute-intensive AI training and deep learning use cases.
Deploy up to 4 resource units per notebook or workout. Storage via Object Storage is unlimited.
Node | Memory | vCore | GPU | Ephemeral local storage | Public network | Price |
A100-1-GPU | 160 GB | 13 | NVIDIA Ampere A100 80GB 19.5 TFLOPS | 80 GB | 1.5 Gbps | $4.2 |
ai1-1-GPU | 40 GB | 13 | NVIDIA Tesla V100S 32GB 16.4 TFLOPS | 750 GB | 1.5 Gbps | $2.702 |
ai1-the-1-GPU | 40 GB | 13 | NVIDIA Tesla V100S 32GB 16.4 TFLOPS | 500 GB | 1.5 Gbps | $1.26 |
H100-1-GPU | 350 GB | 28 | NVIDIA Hopper H100 80GB 51 TFLOPS | 3072 GB | 5 Gbps | $4.34 |
L4-1-GPU | 80 GB | 20 | NVIDIA Ada Lovelace L4 24GB 30.3 TFLOPS | 500 GB | 5 Gbps | $1.162 |
L40S-1-GPU | 80 GB | 13 | NVIDIA Ada Lovelace L40S 48 GB 91.6 TFLOPS | 1536 GB | 5 Gbps | $2.17
|
AI Deploy
CPU Units
Take advantage of guaranteed CPU resources, adapted to your test, pre-production or basic machine learning phases.
Deploy up to 12 resource units per node. Storage via Object Storage is unlimited.
Node | Memory | vCore | Ephemeral local storage | Public network | Price |
ai1-1-CPU | 4 GB | 1 | 40 GB | 500 Mbps | $0.042 |
GPUs
Deploy GPU resources for all your compute-intensive use cases, such as AI training and deep learning.
Deploy up to 4 resource units per node. Storage via Object Storage is unlimited.
Node | Memory | vCore | Ephemeral local storage | Public network | Price |
A100-1-GPU | 160 GB | 13 | 80 GB | 1.5 Gbps | $4.2 |
ai1-1-GPU | 40 GB | 13 | 750 GB | 1.5 Gbps | $2.702 |
ai1-the-1-GPU | 40 GB | 13 | 500 GB | 1.5 Gbps | $1.26 |
H100-1-GPU | 350 GB | 28 | 3072 GB | 5 Gbps | $4.34 |
L4-1-GPU | 80 GB | 20 | 500 GB | 5 Gbps | $1.162 |
L40S-1-GPU | 80 GB | 13 | 1536 GB | 5 Gbps | $2.17
|
Lettria: Licensing by model
Each partner has its own pricing per model and per resource, which can be found below.
For reasons of simplification, the price displayed corresponds to the hourly rate. However, the billing granularity is by the minute.
This pricing is in addition to the cost of configuring AI Deploy physical resources.
Node | Model | Category | Resource Type | Number of API calls | Price |
lettria-emotion-cpu | Emotion Analysis | NLP - Information extraction | CPU | Unlimited | $2.1 |
lettria-emotion-gpu | Emotion Analysis | NLP - Information extraction | GPU | Unlimited | $3.5 |
lettria-named-entity-recognition-cpu | Named Entity Recognition | NLP - Information extraction | CPU | Unlimited | $2.1 |
lettria-named-entity-recognition-gpu | Named Entity Recognition | NLP - Information extraction | GPU | Unlimited | $3.5 |
lettria-sentiment-cpu | Sentiment Analysis | NLP - Information extraction | CPU | Unlimited | $2.1 |
lettria-sentiment-gpu | Sentiment Analysis | NLP - Information extraction | GPU | Unlimited | $3.5 |
Voxist: Pay-as-you-go pricing model
This partner has its own time-of-use pricing per model, which can be found below.
For reasons of simplification, the price displayed corresponds to the hourly rate. However, the billing granularity is by the minute.
This pricing is in addition to the cost of configuring AI Deploy physical resources.
Node | Model | Category | Level | Resource Type | Number of API calls | Price |
cpu-bracket1 | Transcript Standard | NLP - Speech Recognition | From 1 to 100,000 minutes | CPU | Unlimited | $1.232 |
cpu-bracket2 | Transcript Standard | NLP - Speech Recognition | From 100,001 to 500,000 minutes | CPU | Unlimited | $0.938 |
cpu-bracket3 | Transcript Standard | NLP - Speech Recognition | From 500 001 to 1 000 000 minutes | CPU | Unlimited | $0.742 |
cpu-bracket4 | Transcript Standard | NLP - Speech Recognition | From 1,000,001 to 3,000,000 minutes | CPU | Unlimited | $0.546 |
cpu-bracket5 | Transcript Standard | NLP - Speech Recognition | From 3,000,001 to 6,000,000 minutes | CPU | Unlimited | $0.49 |
GPU-Bracket1 | Transcript Standard | NLP - Speech Recognition | From 1 to 100,000 minutes | GPU | Unlimited | $1.232 |
GPU-Bracket2 | Transcript Standard | NLP - Speech Recognition | From 100,001 to 500,000 minutes | GPU | Unlimited | $0.938 |
GPU-Bracket3 | Transcript Standard | NLP - Speech Recognition | From 500 001 to 1 000 000 minutes | GPU | Unlimited | $0.742 |
GPU-Bracket4 | Transcript Standard | NLP - Speech Recognition | From 1,000,001 to 3,000,000 minutes | GPU | Unlimited | $0.546 |
GPU-Bracket5 | Transcript Standard | NLP - Speech Recognition | From 3,000,001 to 6,000,000 minutes | GPU | Unlimited | $0.49 |