Together AI: Scale your AI with fast, optimized model building & deployment.

Llama 4 Maverick

$0.27 per million input tokens and $0.85 per million output tokens.

Llama 3.3 70B Instruct-Turbo

$0.88 per million tokens for both input and output.

DeepSeek-R1

$3.00 per million input tokens and $7.00 per million output tokens.

1x H200 141GB

$4.99 per hour.

1x H100 80GB

$3.36 per hour.

1x A100 SXM 80GB

$2.56 per hour.

Supervised Fine-Tuning (LoRA) Up to 16B parameters

$0.48 per million tokens for up to 16B parameters.

Supervised Fine-Tuning (LoRA) 17B–69B parameters

$1.50 per million tokens for 17B–69B parameters.

Supervised Fine-Tuning (LoRA) 70B–100B parameters

$2.90 per million tokens for 70B–100B parameters.

Full Fine-Tuning Up to 16B parameters

$0.54 per million tokens for up to 16B parameters.

Full Fine-Tuning 17B–69B parameters

$1.65 per million tokens for 17B–69B parameters.

Full Fine-Tuning 70B–100B parameters

$3.20 per million tokens for 70B–100B parameters.

Code Sandbox

$0.0446 per vCPU per hour plus $0.0149 per GiB RAM per hour.

Code Interpreter

$0.03 per session (60 minutes).

Instant Clusters NVIDIA HGX H100 SXM 1 Week–3 Months

$2.20 per hour per GPU for 1 Week–3 Months.

Instant Clusters NVIDIA HGX H100 SXM 1–6 Days

$2.50 per hour per GPU for 1–6 Days.

Instant Clusters NVIDIA HGX H100 SXM Hourly

$2.99 per hour per GPU for Hourly.

Reserved Clusters

Starting at $0.98 per hour, with volume discounts of up to 40%.

Scale Plan

Includes up to 9,000 requests per minute, 5 million tokens per minute, discounts, HIPAA compliance, and a 99% availability SLA.

Enterprise Plan

Offers secure, private deployments with all Scale benefits, plus a 99.9% SLA, custom regions, and priority hardware access.