🔥 Clarifai Reasoning Engine
Benchmarked by Artificial Analysis on GPT-OSS-120B → 544 tokens/sec, 3.6s TTFA, $0.16/M — Faster, Cheaper, Adaptive
👉 Learn More
Contact us
Contact us
Join the Discord
Join the Discord
Start for free
Start for free
Start for free
humburger-icon
menu-close

Are your LLMs too slow? Clarifai them.

Benchmark your models on the world's fastest inference engine with a
free 14-day trial. Our self-optimizing technology is built to accelerate complex reasoning tasks on GPUs.

Talk to us about a 14-day Free Trial
aa-gptoss-2graphs-transparent

Pay As You Go

Explore AI using dedicated deployments, serverless pre-trained models in the cloud, our robust API and low-code UIs.

  • Run models directly or via dedicated compute instance
  • Clarifai Reasoning Engine. Accelerate Agentic AI workloads and large reasoning models.
  • Full platform access
  • Promotional access to Local Runners
  • Up to 100 requests per second
No monthly commitment!
Sign Up

Hybrid-Cloud AI Enterprise

Unlimited SaaS or VPC AI development and production workloads.

  • Unlimited API calls
  • Clarifai’s SaaS or private control plane
  • Multi-cloud and multi-region compute planes with top GPUs
  • Optional air-gapped deployments and private data planes
  • Full model exports, leaderboards
  • Custom rate limits
  • Multiple Organizations
  • Role-based access and Teams
  • Enterprise 99.99% SLAs
  • 24/7 dedicated support
Custom Pricing
Talk to our AI experts

Compare Plans

Compare features and benefits across every plan.

Pay As You Go
Get Pay As You Go
Enterprise
Get Enterprise
Usage & limits
Monthly requests 100,000 Unlimited
Requests per second 100 1000+
SDK & API access
Compute
Deployment types SaaS, Local Dev, Hybrid Cloud (Self-Hosted) SaaS, Local Dev, Hybrid Cloud (Self-Hosted), VPC, On-Prem, Air Gapped
NVIDIA GPUs A10G, L4, L40S, A100 H100, H200, B200
Intel & AMD CPUs
Inference
Pre-trained model access
Batch requests
Realtime bi-directional streaming
GPU fractioning
Scale to zero
Spot instances
Development & training
Custom model training Train & deploy Enterprise AI
Model evaluation
Model upload
Model export
Dataset management
Vector search
Automated data labeling

Your LLMs Too Slow? Clarifai them.

Benchmark your models on the world's fastest inference engine, free for 14 days.
Our self-optimizing technology is built to accelerate complex reasoning tasks on GPUs.

Compare Plans

Compare features and benefits across every plan.

Community
Get Community
Essential
Get Essential
Professional
Get Professional
Hybrid AI Enterprise
Contact Us
Private AI Enterprise
Contact Us
Usage & limits
Monthly requests Limited 30,000 100,000 Unlimited Unlimited
Requests per second 1 15 100 1000+ 1000+
SDK & API access
Compute
Deployment types SaaS, Local Dev + Hybrid Cloud
(Self-Hosted) 		+ Hybrid Cloud
(Self-Hosted) 		+ Hybrid Cloud
(Self-Hosted) 		+ VPC, On-Prem, Air Gapped
NVIDIA GPUs Serverless A10G, L4  + L40S + A100, H100, H200, B200 + A100, H100, H200, B200
Intel & AMD CPUs
Inference
Pre-trained model access
Batch requests
Realtime bi-directional streaming
GPU fractioning
Scale to zero
Spot instances
Development & training
Custom model training Fine-tune Train & deploy Full training Enterprise AI
Model evaluation
Model upload
Model export
Dataset management
Vector search
Automated data labeling

Inference Pricing

Over 500 leading open-source and closed-source language, multimodal, image, code, and embedding models are available for serverless and dedicated inference. 

Large Language Models (per 1M tokens)
GPT OSS 120B
Input $0.09 / Output $0.36
Llama-3_2-3B-Instruct
Input $0.13 / Output $0.63
Qwen-2_5-Coder-7B-Instruct-vllm
Input $0.44 / Output $1.32
MiniCPM3-4B
Input $0.69 / Output $1.10
QwQ-32B-AWQ
Input $1.22 / Output $1.56
DeepSeek-R1-Distill-Qwen-7B
Input $0.21 / Output $0.67
Vision Language Models (per 1M tokens)
MiniCPM-o-2_6-language
Input $0.66 / Output $1.11
Qwen2_5-VL-7B-Instruct
Input $0.44 / Output $1.32
GPT-4o
Input $6.25 / Output $10.00
claude-sonnet-4
Input $3.75 / Output $18.75
Small Image Models
Pre-trained classification model
$0.0012 /request
Pre-trained detection model
$0.002 /request
Pre-trained segmentation model
$0.0032 /request
Custom classification model
$0.0032 /request
Custom detection model
$0.005/request
Custom segmentation model
$0.008 /request
Small Language Models
Pre-trained classification model
$0.0012 /request
Pre-trained NER model
$0.002 /request
Custom classification model
$0.0032 /request
Inputs and Vectors
Input and vector ingest general embeddings
$0.0012 /request
Input and vector ingest detection embeddings
$0.002 /request
Active inputs index (monthly)
$0.001 /input
Input and vector search
$0.0012 /request
Model training
Model training job (single GPU)
$4 /hr
Model training containers (multi-GPU)
Contact us

Dedicated Node Pricing

Only pay for the compute you use, down to the minute

Best-in-class model performance, effortless autoscaling, and blazing fast cold starts mean you get the most out of each GPU, saving money along the way.

Request your GPU cluster

Cloud Provider
AWSAmazon Web Services
  • AWSAmazon Web Services
  • Google Cloud Project NodesGoogle Cloud
  • vultrVultr
Provider Region
Countries Flagus-east-1
  • Countries Flagus-east-1
  • Countries Flagus-west-2
  • Countries Flagus-east-4
  • Countries Flagnew-york
Node name Cloud instance name Price per min
NVIDIA T4 16GB XL g4dn.xlarge $0.011
NVIDIA A10G 24GB XL g5.xlarge $0.021
NVIDIA A10G 24GB 2XL g5.2xlarge $0.0253
NVIDIA L4 24GB XL g6.xlarge $0.0168
NVIDIA L4 24GB 2XL g6.2xlarge $0.0204
NVIDIA L40S 48GB XL g6e.xlarge $0.0388
NVIDIA L40S 192GB 12XL g6e.12xlarge $0.2186
AMD EPYC 7000 M t3a.medium $0.0012
AMD EPYC 7000 L t3a.large $0.0016
AMD EPYC 7000 XL t3a.xlarge $0.0031
AMD EPYC 7000 2XL t3a.2xlarge $0.0063
NVIDIA L4 24GB XL g2-standard-4 $0.0147
NVIDIA L4 24GB 2XL g2-standard-8 $0.0178
NVIDIA L4 24GB 3XL g2-standard-12 $0.0208
NVIDIA L4 24GB 4XL g2-standard-16 $0.0239
NVIDIA L4 24GB 5XL g2-standard-32 $0.0361
NVIDIA A100 80GB XL a2-ultragpu-1g $0.1189
NVIDIA H100 80GB XL a3-highgpu-1g $0.2304
Intel ICL/CSL CPU 8GB n2-standard-2 $0.0023
Intel ICL/CSL CPU 16GB n2-standard-4 $0.0046
Intel ICL/CSL CPU 32GB n2-standard-8 $0.0091
Intel ICL/CSL CPU 64GB n2-standard-16 $0.0182
NVIDIA T4 16GB XL g4dn.xlarge $0.011
NVIDIA A10G 24GB XL g5.xlarge $0.021
NVIDIA A10G 24GB 2XL g5.2xlarge $0.0253
NVIDIA L4 24GB XL g6.xlarge $0.0168
NVIDIA L4 24GB 2XL g6.2xlarge $0.0204
NVIDIA L40S 48GB XL g6e.xlarge $0.0388
NVIDIA L40S 192GB 12XL g6e.12xlarge $0.2186
AMD EPYC 7000 M t3a.medium $0.0012
AMD EPYC 7000 L t3a.large $0.0016
AMD EPYC 7000 XL t3a.xlarge $0.0031
AMD EPYC 7000 2XL t3a.2xlarge $0.0063
NVIDIA A16 M 16GB vcg-a16-6c-64g-16vram $0.0098
NVIDIA L40S XL 48GB vcg-l40s-16c-180g-48vram $0.0348
NVIDIA A100 XL 80GB vcg-a100-12c-120g-80vram $0.0499
Contact us to unlock larger GPU instances or any cloud region

Frequently Asked Questions

How does pricing work?

Each plan provides a monthly credit that can be applied toward various operations, such as model predictions, training, and data storage. If your usage exceeds the included credit, additional charges will apply based on the specific operations performed.

 

What are the maximum operations and inputs stored per package?

Each plan has a maximum number of operations and inputs that can be stored monthly. The Essential Plan allows up to 30,000 operations, while the Professional Plan permits up to 100,000. To exceed these operations, contact sales@clarifai.com for a custom package that includes committed volume discounts.

What if I need to cancel my subscription?

You can cancel your subscription any time by downgrading to the Community plan.

Do you offer Professional Services? How much do you charge?

Clarifai has a team of AI experts that will help you implement AI projects. We can also offer custom development, depending on the project's need.  

Please contact us to share more about your project.

Unlock Peak Model Performance

Let our AI experts help you optimize your models for speed, performance and scale.

Consult an Expert