Together AI logo

Together AI

The AI native cloud.

United States San Francisco, CA Founded 2022 4 GPU models

At a glance

prices checked 3h ago
Cheapest GPU
$1.99/hr
NVIDIA H100 SXM5 80GB
Fastest GPU
$4.65/hr
NVIDIA B200 SXM 180GB
GPU models
4
8 configurations
Certifications
2
SOC 2 Type II, ISO 27001

What's good about Together AI

  • Largest cluster in our data 36,000 GB200s
  • Series C $800M
  • Storage egress Zero fees
  • Named customers Cursor, Cohere, DeepMind, ElevenLabs, Salesforce

Available GPUs

Cheapest first, per GPU per hour.

GPU Configuration Memory Price Region
NVIDIA B200 SXM 180GB
Blackwell
1 GPU 180 GB $4.65/hr - Rent
NVIDIA B300 SXM 262GB
Blackwell Ultra
1 GPU 262.5 GB $5.75/hr - Rent
NVIDIA B200 SXM 180GB
Blackwell
1 GPU Reserved 180 GB $7.99/hr - Rent
NVIDIA B300 SXM 262GB
Blackwell Ultra
1 GPU Reserved 262.5 GB $9.89/hr - Rent
NVIDIA H100 SXM5 80GB
Hopper
1 GPU 80 GB $1.99/hr - Rent
NVIDIA H200 SXM 141GB
Hopper
1 GPU 141 GB $2.99/hr - Rent
NVIDIA H100 SXM5 80GB
Hopper
1 GPU Reserved 80 GB $3.45/hr - Rent
NVIDIA H200 SXM 141GB
Hopper
1 GPU Reserved 141 GB $4.15/hr - Rent

Together AI head to head

Compare any two

About Together AI

Together AI sells the whole stack around open models: serverless and batch inference, provisioned throughput with committed capacity, dedicated model and container inference, fine-tuning, AI-optimised storage with no egress fees, and sandboxed code execution. It also rents raw GPU capacity through Accelerated Compute, scaling from self-serve instant clusters to thousands of GPUs, with on-demand B200s available. Founded in 2022, it raised an $800 million Series C and counts Cursor, Cohere, DeepMind, ElevenLabs, Salesforce, Cognition and Jasper among its named customers. Two Together clusters appear in our datacenter directory, including a planned 36,000 GB200 deployment.

36,000 GB200s Largest cluster in our data
Planned deployment recorded in our datacenter directory.

Pricing options

Pay-as-you-gotoken-basedcommitted capacityhourly GPU

Services

Serverless and batch inferenceprovisioned throughputdedicated inferenceGPU clusters (Accelerated Compute)fine-tuningmanaged storagesandboxes

Compliance and certifications

Independently verified standards governing security, privacy and operations.

Ready to deploy on Together AI?

Visit Together AI

© 2026 Flopper.io - Compare the GPUs Powering AI