Together AI

We were founded in 2022 to provide AI cloud infrastructure. We specialize in AI, cloud computing, large language models, open source, and decentralized computing. We are a team of 201-500 people.

Our GPU cloud platform helps AI engineers and researchers train, fine-tune, and run advanced AI models. It’s built for large-scale production, handling trillions of tokens efficiently.

We offer:

• A Model Library of open-source and specialized models with OpenAI-compatible APIs.
• Inference services using our ATLAS speculator system and Together Inference Engine.
• Fine-Tuning to create task-specific models from open-source options.
• Pre-Training with technologies like our Together Kernel Collection (TKC).
• GPU Clusters for global scaling with hardware such as NVIDIA GB200 NVL72 and GB300 NVL72.

Our customers include SaaS companies like Salesforce, Zoom, and Zomato, and AI startups such as ElevenLabs, Hedra, and Cartesia. We also contribute to open-source AI research and projects including FlashAttention, Mixture of Agents, Dragonfly, Red Pajama Datasets, DeepCoder, Open Deep Research, Flash Decoding, and Open Data Scientist Agent.

We are a research-driven AI company. Our mission is to build technology that benefits society. We champion open-source AI, believing that transparent systems foster innovation and lead to the best results for everyone.

Our core principles for open-source AI are:

• Transparency: You should be able to inspect model training and data.
• No Lock-in: You own your fine-tuned models.
• Privacy: Your data privacy is maintained, whether locally or in our secure cloud.

Our values include:

• Empowering Innovation to help the open-source community.
• Discovering the magic, driven by curiosity to improve lives.
• Doing more with less, as optimizers.
• Being open to any prompt, encouraging exploration.
• Practicing model stewardship, building with the purpose of benefiting society.

We are a team of 201-500, with a research-driven approach to AI. Our culture is built on our core values: transparency, discovery, optimization, openness, and stewardship. We emphasize empowering innovation within the open-source community.

Collaboration is key, and our teams—including engineering, product, research, and sales—work closely together. We offer hybrid work models, though some roles require regular in-office presence in locations like Amsterdam and San Francisco to support our collaborative environment.

We offer a strong benefits package. This includes competitive health, dental, and vision insurance, pre-tax flexible spending accounts, and mental health support.

For income protection and retirement, we provide a 401(k) plan, AD&D insurance, life insurance, and short-term (STD) and long-term (LTD) disability insurance.

Our culture and lifestyle perks include monthly team lunches, flexible time off, team celebrations, and a monthly commuting stipend with pre-tax options. We also provide startup equity, competitive compensation, and hybrid work models. Some roles require two days per week in our Amsterdam or San Francisco offices.