Introducing Jamba, our groundbreaking SSM-Transformer open model!
As the first production-grade model based on Mamba architecture, Jamba achieves an unprecedented 3X throughput and fits 140K context on a single GPU.
🥂Meet Jamba ai21.com/jamba
🔨Build on @huggingface
AI21 Labs
652 posts
AI21 Labs builds Foundation Models and AI Systems for the enterprise that accelerate the use of GenAI in production.
Meet AI21 Maestro
ai21.com
Joined August 2019
- Attention was never enough. The hybrid LLM era is here—and it’s moving fast. From Mamba to Jamba to Bamba, we mapped every major model that’s challenged the Transformer default in the past 18 months. 🧵 A timeline of what’s changed and why it matters ↓ 🔗
- We released the #Jamba 1.5 open model family: - 256K #contextwindow - Up to 2.5X faster on #longcontext in its size class - Native support for structured JSON output, function calling, digesting doc objects & generating citations twtr.to/giIEE #AI #LLM #AI21Jamba
- 📄Jamba-1.5 whitepaper is out! The whitepaper details the architecture, training schemes, novelties and in-depth evaluations of our new long context hybrid SSM-Transformer models - Jamba-1.5-Large and Jamba-1.5-Mini. Arxiv: arxiv.org/abs/2408.12570 Here are some highlights and
- Today we’re launching AI21 Studio, a platform where you can instantly access our state-of-the-art language models to build your own applications - including a 178B parameters model, Jurassic-1 Jumbo. We can’t wait to see what you create!
- We just released Jamba-Instruct! Built from our groundbreaking SSM-Transformer Jamba architecture, Jamba-Instruct brings the same technological innovation to the enterprise via an aligned model. With leading quality benchmarks, a 256K context window, and the most competitive
- 📄Jamba whitepaper is out! The whitepaper details our in-depth ablations on this novel hybrid SSM-Transformer architecture, and how we chose to interleave Mamba, Transformer and MoE. arxiv.org/abs/2403.19887 Here are some highlights from the paper 👇1/6
- 🚀 Introducing Structured RAG (S-RAG) S-RAG transforms unstructured data into a structured, query-aware representation. It then uses formal queries over structured data at runtime, so AI21 Maestro can retrieve accurate values, ensure completeness, handle inconsistencies, and
- 1/5 Releasing Jamba Reasoning 3B under Apache 2.0: Hybrid SSM-Transformer architecture that tops accuracy & speed across record context lengths. e.g. 3-5X faster than Llama 3.2 3B and Qwen3 4B at 32K tokens.
- Today we launched Jamba 1.6, the best open model for private enterprise deployment. AI21’s Jamba outperforms Cohere, Mistral and Llama on key benchmarks, including Arena Hard, and rivals leading closed models while maintaining unmatched speed and quality. Now available on
- Now live. A new update to our Jamba open model family 🎉 Same hybrid SSM-Transformer architecture, 256K context window, efficiency gains & open weights. Now with improved grounding & instruction following. Try it on AI21 Studio or download from @huggingface 🤗 More on what
- We are warning about scams being perpetrated by malicious actors falsely claiming to be AI21-related crypto/tokens. We explicitly clarify: AI21 has absolutely no connection, direct or indirect, to any cryptocurrency or tokens whatsoever. These cases have been reported to X and
- A language model that can talk to APIs? 😮 Say hello to Jurassic-X! 🦕🦕🦕 Read more on our blog, play with the demo, and sign-up for early access
- We're thrilled to announce that we've raised $155 million in series C funding, proudly joining the unicorn club with a valuation of $1.4 billion! 🦄












