AI21 LabsReleased March 28, 2024

Jamba 1.5 Mini (SSM)

Open Source52B (12B active) parameters

Context

256K

Input

/bin/zsh.20

Key Specifications

🏆

Arena Rank

Not disclosed

📐

Context Window

256K

📥

Input Price

per 1M tokens

/bin/zsh.20

📤

Output Price

per 1M tokens

/bin/zsh.40

🧠

Parameters

52B (12B active)

🔓

Open Source

Yes

Best For

Efficient long-context processingthroughput

About Jamba 1.5 Mini (SSM)

Jamba is AI21 Labs' pioneering hybrid architecture that combines Mamba state-space model layers with transformer attention layers. This novel approach achieves 3x throughput improvement over pure transformer models of equivalent quality. The Jamba architecture represents a significant step toward more efficient AI models that can process long contexts without the quadratic cost of full attention.

Built byAI21 Labs

Pricing per 1M tokens

Input Tokens

/bin/zsh.20

Output Tokens

/bin/zsh.40

Frequently Asked Questions

What is Jamba 1.5 Mini (SSM)?
Jamba is AI21 Labs' pioneering hybrid architecture that combines Mamba state-space model layers with transformer attention layers. This novel approach achieves 3x throughput improvement over pure transformer models of equivalent quality. The Jamba architecture represents a significant step toward more efficient AI models that can process long contexts without the quadratic cost of full attention.
How much does Jamba 1.5 Mini (SSM) cost?
Jamba 1.5 Mini (SSM) costs /bin/zsh.20 per 1 million input tokens and /bin/zsh.40 per 1 million output tokens. Pricing is based on token usage, making it cost-effective for both small and large-scale applications.
What is Jamba 1.5 Mini (SSM)'s context window?
Jamba 1.5 Mini (SSM) has a context window of 256K tokens. This determines how much text the model can process in a single request — larger context windows allow the model to handle longer documents, maintain more conversation history, and reason over bigger codebases.
Is Jamba 1.5 Mini (SSM) open source?
Yes, Jamba 1.5 Mini (SSM) is open source. This means the model weights are publicly available, allowing developers and organizations to download, fine-tune, and self-host the model on their own infrastructure. Open-source models offer greater flexibility and data privacy control.
What is Jamba 1.5 Mini (SSM) best for?
Jamba 1.5 Mini (SSM) is best suited for: Efficient long-context processing, throughput. These use cases leverage the model's specific strengths in terms of capability, speed, and cost-effectiveness within AI21 Labs's model lineup.