🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

A nice view into the latent subconscious by echo.hive

.Together AI is possibly the best place to transition to B200 and GB200 with our new infrastructure and deep support for Blackwell kernels and optimization.

Learn how to adapt OSS models to your tasks to achieve better quality / performance.

🥳 Congratulations to our customer Vercept on their launch today! Vercept is reinventing how humans use computers 💻 Vy is a first glimpse at AI that sees and uses your computer just like you do. Vercept built VyUI, an AI model bridging the gap between language and your screen.

Qwen 3 235B now on Together AI API! Qwen 3 is a reasoning model that has a non-reasoning instruct mode with allowance for setting a thinking budget. It’s efficient ($0.20/M input & $0.60/M output on our throughput optimized endpoint) and fantastic on a variety of

Join us this Monday, May 5th at 10:00 AM PT for a talk on Matryoshka Principles for Adaptive Intelligence by Aditya Kusupati, Staff Research Scientist Google DeepMind. Register below 🧵📷🧵!

🚀 Arcee AI moved from AWS to Together Dedicated Endpoints—unlocking simpler operations, massive latency improvements, and greater cost-efficiency for their specialized small language models. Full migration story below 👇

There’s a lot of FUD around safety of OSS models, but these models allow us to study safety implications of AI in an open and transparent way. Excited for the partnership with General Analysis to comprehensively stress test OSS (and closed) models.

Congratulations to our portfolio companies, Together AI and Refuel, on uniting their strengths to power the next generation of AI infrastructure! Together AI’s AI Acceleration Cloud enables developers and enterprises to train and deploy generative AI models with speed,

A new effort led by Percy Liang to build open models in a radically participatory way. The project is now live, including the first Marin models. Here's the project website: marin.community And the 8B model is live on Together's model platform:

Together Code SandBox and Code Interpreter, the fastest VM technology on the planet, designed for AI model generated code execution, text-to-app platforms and RL pipelines, GAs today! Together AI

The new R1 is live on Together AI API as well as chat.together.ai!

What’s a good way to estimate the number of bits created daily from human effort?

.Together AI API has the fastest DeepSeek v3 endpoint (2x faster than next best API endpoint) and almost 5x faster than DeepSeek API. See how to use it directly with Cline to make all your Cline workflows snappier!

Bulk intelligence now available Together AI. Our new Batch API offers broad access to SOTA open source LLMs for high throughput uses. This is great for synthetic data generation, benchmarking, content review and summarization, document extraction and more. Also intro

It's wild that within the next 12 months we are going to witness the end of hand-written code.

.Together AI is building 2 gigawatts of AI factories (~100,000 GPUs) in the EU over the next 4 years with the first phase live in H2 '2025. AI compute is at <1% saturation relative to our 2035 forecast and we are starting early to build a large-scale sustainable AI cloud

One of the cool innovations in DeepSeek's sparse and wide expert parallel architecture is that experts can be split across many 8xH100 servers. We've took DeepSeek's starting point and developed a new inference engine for highly parallel inference and it's live today for R1.