From $0.01 usage-based · Free trial🇺🇸
US inference platform for open, fine-tuned and custom models on autoscaling GPUs, with Model APIs, training and self-hosting.
Together AI AI features and pricing (October 2026)
AI cloud for serverless inference of open models, dedicated endpoints, fine-tuning and GPU clusters.
Usage-based
from $0.09 usage-based
Pricing details
Serverless inference per million tokens, for example $0.09 input and $0.28 output for a small model on the pricing page; dedicated H100 from $3.99 per hour (promotional rate), H100 GPU clusters from $3.99 per GPU hour, as of September 2026.
What is Together AI?
Together AI is a cloud platform for running and adapting open AI models. Development teams call models through serverless APIs, fine-tune them or rent dedicated GPUs and clusters.
Together AI groups its offer into inference services (serverless, batch, provisioned throughput, dedicated models and containers), compute (GPU clusters from self-service to thousands of GPUs, storage, sandboxes) and model optimization through fine-tuning. The company was founded in 2022, is based in San Francisco and runs a research group known for work on FlashAttention. It names Zoom, Cursor and Salesforce among its users.
Together AI AI features and use cases
- Calls open models by APIServerless inference gives developers access to a catalog of open models with pay-per-token billing. Product teams can compare models in an application without running their own GPUs.
- Adapts models to company dataFine-tuning jobs use supervised fine-tuning or direct preference optimization on open models, billed per million training tokens. The adapted model can then run on a dedicated endpoint.
- Keeps prompts out of storageA Zero Data Retention setting prevents prompts and outputs from being stored or used for training. This helps teams that process confidential text through the API.
What sets Together AI apart
- Covers the path from serverless API calls to fine-tuning, dedicated endpoints and multi-thousand GPU clusters in one account.
- Publishes token, GPU and fine-tuning prices, with a Zero Data Retention option in account settings.
Where Together AI is strongest
United States Available internationally
Together Computer, Inc. is based in San Francisco, California, according to its terms of service.
Questions to ask before you buy Together AI
Check where requests are processed, since the privacy policy names no data center locations and relies on standard contractual clauses for EU data. Activate Zero Data Retention if prompts contain personal data. Token costs scale with usage, so set budget alerts. A data processing agreement is not linked publicly and should be requested.
Together AI reviews
No reviews yet. Be the first to share how your team uses it.
Every review shows its verification level. We read each one before it goes live. How we check reviews
Questions about Together AI
How does Together AI charge?
Serverless models are billed per million input and output tokens, dedicated endpoints and clusters per GPU hour, and fine-tuning per million training tokens with a minimum per job.
Does Together AI train on my data?
According to its privacy policy, not without explicit consent. A Zero Data Retention option also prevents storage of prompts and outputs.
Can I rent GPU clusters?
Yes. On-demand clusters start at $3.99 per H100 GPU hour, and reserved capacity over 7 to 180 days or more brings discounts.
Together AI update historyno updates yet
No changes recorded for Together AI yet.
Changes the vendor made and confirmed user reports, each checked against the vendor's own pages. We review every tool monthly.
Alternatives to Together AI
Tools in the same category (Serverless inference) with the most overlapping markets.
From $0.02 usage-based🇺🇸
Inference API for more than 100 open and commercial models billed per token, plus GPU instances and dedicated clusters.
From $250 per month · Free plan🇺🇸
Serverless Python platform for AI inference, training, batch jobs and sandboxes with per-second GPU billing.
Usage-based🇺🇸
GPU cloud with low-cost hourly Pods, serverless endpoints that scale to zero and clusters for developers and small teams.