Together AI

Fine-Tuning to shape open-source models with user data

Evaluations to measure model quality

Managed Storage to store model weights and data securely

GPU Clusters supporting GB300, GB200, B200, H200, and H100 hardware

Supports open-source models including MiniMax M3, Gemma 4 31B, DeepSeek V4 Pro, GLM-5.2, kimi K2.7 Code, and gpt-oss-120B

Provides open-source demo apps and practical implementation cookbooks

Supports building voice agents for production

Serverless inference delivered as high-performance APIs

Batch inference for batch workloads

Provisioned throughput offering token-based capacity with SLAs

Dedicated model inference on custom hardware

Dedicated container inference for custom models

Reliable GPU clusters available at scale

AI Factory for custom infrastructure at frontier scale

Fine-tuning to shape models with customer data

Managed storage to store model weights and data securely

Evaluations to measure model quality

Transparent, flexible pricing across serverless inference, dedicated endpoints, fine-tuning, and GPU clusters, with a free start and on-demand scaling

Together AI Batch API processes thousands of LLM requests at 50% lower cost

Together AI announced its Series C

Together AI and Y Combinator announced a partnership to deliver the first dedicated YC GPU cluster

On-demand B200 GPUs are available on Together GPU Clusters

Together AI now serves MiniMax-M3 for efficient inference

Serverless Inference delivers high-performance inference as APIs

Batch Inference is offered for batch workloads

Provisioned Throughput provides token-based capacity with SLAs

Dedicated Model Inference runs inference on custom hardware

Dedicated Container Inference is offered for custom models

Fine-Tuning allows users to shape models with their data

Supported GPU types include GB300, GB200, B200, H200, and H100

Serverless Inference offering high-performance inference as APIs

Batch Inference designed for batch workloads