Next Era of
On-Device AI Agents

We are building small but powerful on-device multimodal models for super AI agents --- fast, accurate, energy-efficient

Introducing Octopus V3

smallest, most powerful on-device multimodal model for super AI agents

Compact size

Less than 1B parameters

Multimodal

Processes both text and images for function calling

High Performance

On par with a combination of GPT-4V and GPT-4

Multilingual

Fluent in English and Mandarin

Octopus V3 Demo Video

Vision

Nexa AI envisions a world where everyone has access to secure, eco-friendly, and ultra responsive AI—transforming tasks into immediate, actionable solutions right at your fingertips

Octopus V2 Highlights

On-device model

2 billion parameter language model developed for high-performance function calling on edge devices.

Functional tokens

Uses a novel "functional token" strategy to reduced context length by 95%.

Blazing fast

Demonstrated 36x faster inference speed compared to RAG solution, and 168% faster than GPT-4-turbo.

Highly accurate

Achieved 98%+ function call accuracy, surpassing previous models and matching the performance of GPT-4.

Octopus V2 Demo Video

What they're saying

Rowan Cheung
Rundown AI, Founder

“a groundbreaking new framework for on-device AI agents. The new era of on-device AI agents is coming.”

Omar Sanseviero
Hugging face, CLO

“Extremely fast, better than Llama+RAG, great results”

AK
Gradio, ML

“These models possess the crucial ability to call functions, which is essential in creating”

Philipp Schmid
Hugging face, Tech lead & LLMs

“Interesting idea to incorporate the functions into the model with fine-tuning to get reliable generation from small LLMs.”

Aran Komatsuzaki
Teraflop AI, Founder

“For all things tech, Techware is my ultimate destination. Quality, range, and service—impeccable.”

Brett Adcock
Figure AI, Founder

“With just 100 training samples, the model achieved 98% accuracy in selecting the right function, surpassing GPT-4.”

Rowan Cheung
Rundown AI, Founder

“a groundbreaking new framework for on-device AI agents. The new era of on-device AI agents is coming.”

Omar Sanseviero
Hugging face, CLO

“Extremely fast, better than Llama+RAG, great results”

AK
Gradio, ML

“These models possess the crucial ability to call functions, which is essential in creating”

Philipp Schmid
Hugging face, Tech lead & LLMs

“Interesting idea to incorporate the functions into the model with fine-tuning to get reliable generation from small LLMs.”

Aran Komatsuzaki
Teraflop AI, Founder

“For all things tech, Techware is my ultimate destination. Quality, range, and service—impeccable.”

Brett Adcock
Figure AI, Founder

“With just 100 training samples, the model achieved 98% accuracy in selecting the right function, surpassing GPT-4.”

Tom Zschach
SWIFT, CIO

“a groundbreaking new framework for on-device AI agents.”

Julien Chaumond
Hugging Face, CTO
Fredy Del Vecchio
Birdiefy AI, ex CPO& Cofounder

“A monumental leap in function calling efficiency on devices, making real-world applications faster and smarter than ever imagined.”

Raphaël MANSUY
ELITIZON Ltd, CTO

“Octopus v2 represents a major leap towards making powerful AI accessible to everyone.”

Blake Tindol
Stryker, Data Scientist

“Octopus v2 showcases the potential to revolutionize how we interact with technology, emphasizing efficiency and privacy.”

Manoj Kumar
OPPO, Leading Edge AI Team

“A Novel method enabling on-device models with 2 billion parameters to outperform GPT-4 in accuracy and latency, reducing context length by 95%.”

Jon Salisbury
Nexigen, CEO

“I envisioned this last year and it’s happening. Super interesting. So fast.”

Altaf Rehmani
HSBC, Digital Solution Architect

“This is amazing and will pave the path for agents on edge devices.”

Julien Chaumond
Hugging Face, CTO
Naqqash Abbassi
mydost.ai , CTO

“This 2B LLMs is a breakthrough in the application of LLMs for function calling, specifically tailored for Android APIs.”

Fredy Del Vecchio
Birdiefy AI, ex CPO& Cofounder

“A monumental leap in function calling efficiency on devices, making real-world applications faster and smarter than ever imagined.”

Raphaël MANSUY
ELITIZON Ltd, CTO

“Octopus v2 represents a major leap towards making powerful AI accessible to everyone.”

Blake Tindol
Stryker, Data Scientist

“Octopus v2 showcases the potential to revolutionize how we interact with technology, emphasizing efficiency and privacy.”

Manoj Kumar
OPPO, Leading Edge AI Team

“A Novel method enabling on-device models with 2 billion parameters to outperform GPT-4 in accuracy and latency, reducing context length by 95%.”

Jon Salisbury
Nexigen, CEO

“I envisioned this last year and it’s happening. Super interesting. So fast.”

Altaf Rehmani
HSBC, Digital Solution Architect

“This is amazing and will pave the path for agents on edge devices.”

Explore our collection of 200+ Premium Webflow Templates