Sold by: Together AI
Train, fine-tune, and run generative AI models faster, at a lower cost, and at a production scale.
4.3
Overview
Together AI makes it easy to run, fine-tune, and manage open-source and custom models, at the production scale. Our platform enables you to:
Run serverless or on-demand models for more consistent traffic
Scale with monthly reserved instances and VPC for larger deployments
Fine-tune via API and deploy your fine-tuned model for inference
Deploy the Together Enterprise Platform in your VPC on EKS with 2-3x faster inference and up to 50% GPU savings
Manage, orchestrate, and optimize models in one place, achieving low latency and high accuracy for your use cases
What to expect after your purchase:
Once you complete your AWS Marketplace purchase, you will be directed to a form to share your account information. You will also be prompted to set up your account here: http://api.together.ai/ .
Once these steps are complete, our team will be in touch within 48 hours to load the purchased Together AI credits into the account you have created.
For custom pricing & units, reach out to the support team at https://www.together.ai/contact
Highlights
- Run over 200 open-source and custom models available via autoscaling Serverless and Dedicated Instances, with private deployment options for enterprise security.
- Fine-tune generative AI models with your proprietary data and maintain ownership of your custom models.
- Enterprise-grade security and reliability with SOC 2, HIPAA compliance, and custom SLAs.
Details
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
Pricing is based on the duration and terms of your contract with the vendor, and additional usage. You pay upfront or in installments according to your contract terms with the vendor. This entitles you to a specified quantity of use for the contract duration. Usage-based pricing is in effect for overages or additional usage not covered in the contract. These charges are applied on top of the contract price. If you choose not to renew or replace your contract before the contract end date, access to your entitlements will expire.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimension | Description | Cost/month |
|---|---|---|
10,000 Credits | $10,000 Together AI Consumption Units package | $10,000.00 |
25,000 Credits | $25,000 Together AI Consumption Units package | $25,000.00 |
50,000 Credits | $50,000 Together AI Consumption Units package | $50,000.00 |
Dimension | Description | Cost/unit |
|---|---|---|
additional_usage | Additional Credits | $1.00 |
Dimensions summary
You buy prepaid credits in fixed packages: 10,000, 25,000, or 50,000 Consumption Units. Each package is a lump-sum contract purchase, not a recurring subscription. The three packages differ only in the amount of credits included, so you pick the size that matches your expected usage. Credits are drawn down as you consume the platform's services. If you use more than your package covers, the additional usage dimension lets you add more credits. This gives you a base commitment plus an overflow option to handle demand beyond your chosen package.
Top-of-mind questions for buyers
Credits draw down as you use the platform's services. These include running open-source models through a managed inference API, reserving dedicated throughput capacity, deploying models on isolated GPUs, and fine-tuning open models. Usage across text, image, video, code, and voice all consumes credits from your prepaid package.
Consumption is metered by the underlying service you use. Serverless inference is billed per million tokens. Provisioned throughput uses PTUs, a normalized throughput unit. Dedicated inference is billed per GPU-hour. Fine-tuning is billed per million tokens processed. Each service draws credits at its own rate.
The additional usage dimension lets you add more credits when consumption passes your package amount. This works as an overflow so demand beyond your chosen 10,000, 25,000, or 50,000 package can still be served. You buy the additional credits rather than being cut off.
together.ai+2
Helpful?
Vendor refund policy
All fees are non-refundable and non-cancellable except as required by law.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Software as a Service (SaaS)
SaaS delivers cloud-based software applications directly to customers over the internet. You can access these applications through a subscription model. You will pay recurring monthly usage fees through your AWS bill, while AWS handles deployment and infrastructure management, ensuring scalability, reliability, and seamless integration with other AWS services.
Resources
Vendor resources
Support
Vendor support
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
Healthy Together is a health technology company that provides SaaS solutions for Department of Health, Health & Humans Services and Public Health Departments. Their mission is to improve collective health and make government more efficient.
A 5-minute diagnostic that scores your AWS partner readiness across strategy, technical fit, and go-to-market — with a personalised gap analysis and action plan.
VizIQ is a conversational and governed enterprise AI workspace. Business teams connect their data, ask questions in plain language, and create and share governed answers, charts and the SQL behind them, without writing SQL. Every answer resolves against a semantic layer and is logged in an audit trail.
A hands-on 2-day workshop where your business and engineering teams apply the AWS AI-Driven Development Lifecycle (AI-DLC) to a real project, from business intent to a first increment deployed on AWS.
Customer reviews
Sidharth K.
fast, developer-friendly AI model API—but browsing and support need work
Reviewed on Oct 18, 2026
Review provided by G2
What do you like best about the product?
I like the simple UI/UX of the webapp and the speed of model inference and the number of open-source AI models I can access via one API. As a software engineer, I believe that their API integration was highly useful to try out the models and build AI apps. The pricing was also affordable for experimentation and developer-friendly documentation made implementation easier.
What do you dislike about the product?
Sometimes the platform is difficult to browse to compare the models, set up the advanced parameters. Pricing is very reasonable, but can be a little expensive if you run a lot. I would like to have more detailed troubleshooting documents and quicker support for complicated integrations.
What problems is the product solving and how is that benefiting you?
Together AI makes it easy to deploy and connect to large language models without having to run all the infrastructure yourself. I can use it to prototype AI features, experiment with different models, and benchmark inference performance via APIs. This saves on my infrastructure burden and allows my team to focus on application logic rather than infrastructure.
Sanyi S.
Easy model testing and LangChain integration with room to improve
Reviewed on Oct 18, 2026
Review provided by G2
What do you like best about the product?
I use Together AI to test open source language models, integrate inference APIs, and compare outputs before selecting a model for application development. The API documentation is useful, and the playground helps me evaluate responses quickly. I also appreciate the model variety, competitive pricing options, and straightforward integration with frameworks like LangChain. Performance is generally suitable for prototyping and development workflows.
What do you dislike about the product?
The UI/UX could make model comparisons, usage tracking, and configuration options easier to navigate. Response latency can vary depending on the model and workload, so performance testing is still necessary. Support documentation could also provide more detailed troubleshooting examples for integration issues.
What problems is the product solving and how is that benefiting you?
Together AI helps me build and test AI-powered applications without managing the entire model infrastructure myself. I can experiment with different language models, compare AI-generated responses, and integrate inference APIs into my development workflow. This reduces setup time, speeds up prototyping, and helps me identify suitable models for specific tasks while keeping development costs under control.
Ramesh S.
A Powerful Hub for Comparing AI Models Across Real-World Use Cases
Reviewed on Oct 18, 2026
Review provided by G2
What do you like best about the product?
Together.ai provides us with a single service to compare multiple AI models across different use cases. It's an extremely helpful way for me to explore how a model might perform on use cases such as document analysis, information extraction, and automation of work processes. Comparing outputs from multiple models helps me identify potential solutions and provides our technical team with a starting point.
What do you dislike about the product?
Deciding on the best model may be an iterative process since the response quality, speed, and price all differ based on the use case. I want to see more business-oriented comparisons and examples of success cases so we can help our teams determine the best fit for various enterprise use cases without having to spend too much time experimenting.
What problems is the product solving and how is that benefiting you?
Turning an AI solution to a business process can demand plenty of investigation and experimentation. Together.ai provides your teams with many models to test against similar tasks, such as extracting data from documents. This will help provide more clear requirements analysis, more opportunity to automate, and a much better conversation between business experts and IT developers.
Rukhsana N.
Single API Access to Powerful AI Models That Accelerates Application Development
Reviewed on Oct 18, 2026
Review provided by G2
What do you like best about the product?
The breadth of open source models available through a single clean API has completely changed how fast our team can prototype and ship AI powered features because instead of spending time evaluating and deploying models ourselves we can test different options through one consistent interface and find what works best within hours and the inference speed has been impressive enough that we have been comfortable building user facing features directly on top of it without worrying about latency impacting the experience and what caught us off guard was how well the API handles concurrent requests at scale because it has held up reliably even as our user traffic has grown significantly.
What do you dislike about the product?
Fine tuning documentation for more specialized application use cases could be more detailed and practical and having more code focused examples and sample applications in the documentation would speed up development significantly for teams building production AI features for the first time.
What problems is the product solving and how is that benefiting you?
We used to struggle with the complexity of managing our own model inference infrastructure which created a significant barrier every time we wanted to add a new AI capability to our application but now we can integrate new models through a single API call which has reduced our AI feature development time by nearly 60 percent and the cost efficiency means we can experiment with more ideas without worrying about infrastructure costs spinning out of control during the prototyping phase.
megha s.
Efficient AI model testing, but cost visibility and support need improvement
Reviewed on Oct 18, 2026
Review provided by G2
What do you like best about the product?
Together AI assists me in trying out various language models and testing their efficacy for my software development projects. The ability to access the models through an API is really handy for trying out features and prompts and comparing outcomes for the various models, rather than having to develop separate integrations for each test. It provides real-time flexibility to my engineering work.
What do you dislike about the product?
Different models and workloads impact response time and how the model performs. I would like more clear answers on usage-cost, easier-to-understand monitoring tools, and more consistent troubleshooting guides. Engineers would also benefit from better integration examples and quicker responses from support to troubleshoot deployment issues.
What problems is the product solving and how is that benefiting you?
Together AI enables me to try different models of language and see how well they work for my software development use cases. The API is convenient for rapid iteration, to try out prompts, and to compare outputs from different models without creating individual integrations for each experiment. The platform supports engineering workflows, though the UI/UX allows for more convenient selection and management of models.