PrismML has successfully optimized its tiny large language models to run locally on Qualcomm-powered smart glasses, signaling a shift toward private, low-latency wearable AI.
#startups #PrismML #Qualcomm
#startups #PrismML #Qualcomm
Edge AI Evolves: PrismML Brings Local Intelligence to the Wearable Frontier
6ic.com
September 25, 2026 at 3:15 PM
PrismML brings its tiny LLMs to Qualcomm-powered smart glasses
PrismML brings its tiny LLMs to Qualcomm-powered smart glasses
Prism's larger goal is open-weight AI that runs on devices and makes better use of the computing power they already have.
techcrunch.com
September 24, 2026 at 7:04 PM
PrismML brings its tiny LLMs to Qualcomm-powered smart glasses
This is insane if true. An 8b parameter open model running with only 1.15 GB of RAM. Told you crazy efficiency improvements were coming down the pipe
Today, PrismML is emerging from stealth. We are an AI lab with Caltech origins centered on building the most concentrated form of intelligence.
At PrismML, we believe that the next major leaps in AI…...
Today, PrismML is emerging from stealth. We are an AI lab with Caltech origins centered on building the most concentrated form of intelligence.
At PrismML, we believe that the next major leaps in AI ...
www.linkedin.com
April 1, 2026 at 10:40 AM
This is insane if true. An 8b parameter open model running with only 1.15 GB of RAM. Told you crazy efficiency improvements were coming down the pipe
PrismML hopes its tiny LLM could change how we all use AI
PrismML hopes its tiny LLM could change how we all use AI
If AI lab PrismML isn't on your radar yet, it should be.
techcrunch.com
September 17, 2026 at 10:35 PM
PrismML hopes its tiny LLM could change how we all use AI
PrismML releases 1-bit LLM (open-weight), or a 8B LLM that fits in 1.15GM of VRAM
Website: prismml.com
Blog: prismml.com/news/bonsai-8b
HuggingFace: huggingface.co/collections/...
Website: prismml.com
Blog: prismml.com/news/bonsai-8b
HuggingFace: huggingface.co/collections/...
March 31, 2026 at 11:44 PM
PrismML releases 1-bit LLM (open-weight), or a 8B LLM that fits in 1.15GM of VRAM
Website: prismml.com
Blog: prismml.com/news/bonsai-8b
HuggingFace: huggingface.co/collections/...
Website: prismml.com
Blog: prismml.com/news/bonsai-8b
HuggingFace: huggingface.co/collections/...
Bonsai 2 27b! 🌳
prismml.com/news/bonsai-...
x.com/PrismML/stat...
5.9gb compressed version of Qwen 3.8 27b while retaining 98.2% of aggregate benchmark performance
prismml.com/news/bonsai-...
x.com/PrismML/stat...
5.9gb compressed version of Qwen 3.8 27b while retaining 98.2% of aggregate benchmark performance
September 17, 2026 at 9:34 PM
Bonsai 2 27b! 🌳
prismml.com/news/bonsai-...
x.com/PrismML/stat...
5.9gb compressed version of Qwen 3.8 27b while retaining 98.2% of aggregate benchmark performance
prismml.com/news/bonsai-...
x.com/PrismML/stat...
5.9gb compressed version of Qwen 3.8 27b while retaining 98.2% of aggregate benchmark performance
‘The technology, according to PrismML, could ultimately extend well beyond phones and laptops to robotics, autonomous systems and other products that need to make decisions quickly without relying on a cloud connection.’ www.cnbc.com/2026/07/14/a...
Apple in talks with startup that shrinks AI models to run on an iPhone
PrismML says its compressed version of Alibaba’s Qwen model uses up to 15 times less memory, potentially advancing Apple’s AI push.
www.cnbc.com
July 14, 2026 at 5:06 PM
‘The technology, according to PrismML, could ultimately extend well beyond phones and laptops to robotics, autonomous systems and other products that need to make decisions quickly without relying on a cloud connection.’ www.cnbc.com/2026/07/14/a...
Blog: prismml.com/news/bonsai-...
Whitepaper: github.com/PrismML-Eng/...
Models: huggingface.co/collections/...
WebGPU Demo: huggingface.co/spaces/webml...
GitHub: github.com/PrismML-Eng/...
Docs: docs.prismml.com
Whitepaper: github.com/PrismML-Eng/...
Models: huggingface.co/collections/...
WebGPU Demo: huggingface.co/spaces/webml...
GitHub: github.com/PrismML-Eng/...
Docs: docs.prismml.com
September 18, 2026 at 1:21 PM
Blog: prismml.com/news/bonsai-...
Whitepaper: github.com/PrismML-Eng/...
Models: huggingface.co/collections/...
WebGPU Demo: huggingface.co/spaces/webml...
GitHub: github.com/PrismML-Eng/...
Docs: docs.prismml.com
Whitepaper: github.com/PrismML-Eng/...
Models: huggingface.co/collections/...
WebGPU Demo: huggingface.co/spaces/webml...
GitHub: github.com/PrismML-Eng/...
Docs: docs.prismml.com
ah! don't forget the whitepaper!
github.com/PrismML-Eng/...
github.com/PrismML-Eng/...
Bonsai-demo/bonsai-2-27b-whitepaper.pdf at main · PrismML-Eng/Bonsai-demo
Bonsai Demo. Contribute to PrismML-Eng/Bonsai-demo development by creating an account on GitHub.
github.com
September 17, 2026 at 10:28 PM
ah! don't forget the whitepaper!
github.com/PrismML-Eng/...
github.com/PrismML-Eng/...
Blog: prismml.com/news/ternary...
Whitepaper: github.com/PrismML-Eng/...
Models: huggingface.co/collections/...
HuggingFace Demo: huggingface.co/spaces/webml...
GitHub: github.com/PrismML-Eng/...
Whitepaper: github.com/PrismML-Eng/...
Models: huggingface.co/collections/...
HuggingFace Demo: huggingface.co/spaces/webml...
GitHub: github.com/PrismML-Eng/...
April 17, 2026 at 4:57 AM
Blog: prismml.com/news/ternary...
Whitepaper: github.com/PrismML-Eng/...
Models: huggingface.co/collections/...
HuggingFace Demo: huggingface.co/spaces/webml...
GitHub: github.com/PrismML-Eng/...
Whitepaper: github.com/PrismML-Eng/...
Models: huggingface.co/collections/...
HuggingFace Demo: huggingface.co/spaces/webml...
GitHub: github.com/PrismML-Eng/...
PrismML — PrismML Brings 1-Bit Bonsai Models to AI Smart Glasses Powered by Snapdragon prismml.com/news/prismml...
PrismML — PrismML Brings 1-Bit Bonsai Models to AI Smart Glasses Powered by Snapdragon
1-bit Bonsai dramatically increases the on-device intelligence of smart glasses, fitting 4× larger models onto the same memory footprint
prismml.com
September 27, 2026 at 10:20 PM
PrismML — PrismML Brings 1-Bit Bonsai Models to AI Smart Glasses Powered by Snapdragon prismml.com/news/prismml...
Bonsai-2 てのが出ているみたい
後で試す
PrismML-Eng/Bonsai-demo: Bonsai Demo https://github.com/PrismML-Eng/Bonsai-demo
後で試す
PrismML-Eng/Bonsai-demo: Bonsai Demo https://github.com/PrismML-Eng/Bonsai-demo
GitHub - PrismML-Eng/Bonsai-demo: Bonsai Demo
Bonsai Demo. Contribute to PrismML-Eng/Bonsai-demo development by creating an account on GitHub.
github.com
September 27, 2026 at 10:35 PM
Bonsai-2 てのが出ているみたい
後で試す
PrismML-Eng/Bonsai-demo: Bonsai Demo https://github.com/PrismML-Eng/Bonsai-demo
後で試す
PrismML-Eng/Bonsai-demo: Bonsai Demo https://github.com/PrismML-Eng/Bonsai-demo
PrismML выводит свои миниатюрные LLM на умные очки на базе Qualcomm
Более масштабная цель Prism — это открытый ИИ, который работает на устройствах и более эффективно использует уже имеющуюся у них вычислительную мощность.
Telegram ИИ Дайджест
#ai #llm #ml
Более масштабная цель Prism — это открытый ИИ, который работает на устройствах и более эффективно использует уже имеющуюся у них вычислительную мощность.
Telegram ИИ Дайджест
#ai #llm #ml
PrismML brings its tiny LLMs to Qualcomm-powered smart glasses
techcrunch.com
September 25, 2026 at 6:52 PM
PrismML выводит свои миниатюрные LLM на умные очки на базе Qualcomm
Более масштабная цель Prism — это открытый ИИ, который работает на устройствах и более эффективно использует уже имеющуюся у них вычислительную мощность.
Telegram ИИ Дайджест
#ai #llm #ml
Более масштабная цель Prism — это открытый ИИ, который работает на устройствах и более эффективно использует уже имеющуюся у них вычислительную мощность.
Telegram ИИ Дайджест
#ai #llm #ml
Read the full announcement: https://prismml.com/news/prismml-brings-1-bit-bonsai-models-to-ai-smart-glasses-powered-by-snapdragon
PrismML — PrismML Brings 1-Bit Bonsai Models to AI Smart Glasses Powered by Snapdragon
1-bit Bonsai dramatically increases the on-device intelligence of smart glasses, fitting 4× larger models onto the same memory footprint
prismml.com
September 23, 2026 at 8:38 PM
Read the full announcement: https://prismml.com/news/prismml-brings-1-bit-bonsai-models-to-ai-smart-glasses-powered-by-snapdragon
PrismML debuts energy-sipping 1-bit LLM in bid to free AI from the cloud
PrismML debuts energy-sipping 1-bit LLM in bid to free AI from the cloud
Bonasi 8B model is competitive with other 8B models but 14x smaller and 5x more energy efficient
PrismML, an AI venture out of Caltech, has released a 1-bit large language model that outperforms weightier models, with the expectation that it will improve AI efficiency and viability on mobile devices, among other applications.…
go.theregister.com
April 5, 2026 at 8:14 AM
PrismML debuts energy-sipping 1-bit LLM in bid to free AI from the cloud
PrismML released Ternary Bonsai: 1.58 bit LLM (open-weight)
Using ternary weights {-1, 0, +1}, they built a family of models that are 9x smaller than their 16-bit counterparts while outperforming most models in their respective parameter classes on standard benchmarks.
Using ternary weights {-1, 0, +1}, they built a family of models that are 9x smaller than their 16-bit counterparts while outperforming most models in their respective parameter classes on standard benchmarks.
April 17, 2026 at 4:57 AM
PrismML released Ternary Bonsai: 1.58 bit LLM (open-weight)
Using ternary weights {-1, 0, +1}, they built a family of models that are 9x smaller than their 16-bit counterparts while outperforming most models in their respective parameter classes on standard benchmarks.
Using ternary weights {-1, 0, +1}, they built a family of models that are 9x smaller than their 16-bit counterparts while outperforming most models in their respective parameter classes on standard benchmarks.
Five days after initial reports that there was an AI-model shrinking technology by PrismML suitable for iPhones, the company has confirmed it is talking to Apple over the use of its technology.
PrismML confirms it is in talks with Apple about AI model-shrinking tech
Five days after initial reports that there was an AI-model shrinking technology by PrismML suitable for iPhones, the company has confirmed it is talking to Apple over the use of its technology.
appleinsider.com
July 14, 2026 at 5:32 PM
Five days after initial reports that there was an AI-model shrinking technology by PrismML suitable for iPhones, the company has confirmed it is talking to Apple over the use of its technology.
prismml.com New neolab with a cool logo and a focus on on-device models. Their first release is a *1 BIT* model competitive with similar models around the same size (in GB, not parameters) and pushes the Pareto frontier for size/performance
PrismML — Concentrating intelligence
Large models can't fit on smartphones. Datacenters can't sustain them. PrismML is building ultra dense intelligence to solve both.
prismml.com
March 31, 2026 at 8:08 PM
prismml.com New neolab with a cool logo and a focus on on-device models. Their first release is a *1 BIT* model competitive with similar models around the same size (in GB, not parameters) and pushes the Pareto frontier for size/performance
Code: https://github.com/PrismML-Eng/Bonsai-demo/tree/agent-demo-bonsaios https://twitter.com/PrismML/status/2103259930614767643
Video: https://twitter.com/prismml/status/2103259930614767643 (2/2)
Video: https://twitter.com/prismml/status/2103259930614767643 (2/2)
September 24, 2026 at 11:08 PM
What are the best AI models you can fit on disks? Here are my guesses:
CD (700 MB): PrismML Bonsai 4B (1-bit quant, trained from Qwen)
DVD (4.7 GB): Qwen 3.5 9B quantized to 3 bits/parameter
Blu-ray (25 GB): Gemma 4 31B quantized to 5 bits/parameter
CD (700 MB): PrismML Bonsai 4B (1-bit quant, trained from Qwen)
DVD (4.7 GB): Qwen 3.5 9B quantized to 3 bits/parameter
Blu-ray (25 GB): Gemma 4 31B quantized to 5 bits/parameter
April 12, 2026 at 6:09 PM
What are the best AI models you can fit on disks? Here are my guesses:
CD (700 MB): PrismML Bonsai 4B (1-bit quant, trained from Qwen)
DVD (4.7 GB): Qwen 3.5 9B quantized to 3 bits/parameter
Blu-ray (25 GB): Gemma 4 31B quantized to 5 bits/parameter
CD (700 MB): PrismML Bonsai 4B (1-bit quant, trained from Qwen)
DVD (4.7 GB): Qwen 3.5 9B quantized to 3 bits/parameter
Blu-ray (25 GB): Gemma 4 31B quantized to 5 bits/parameter
PrismML's tiny LLMs hit smart glasses
PrismML has developed its compact 2B parameter LLMs for Qualcomm Snapdragon-powered smart glasses, shrinking larger models 4x while keeping performance.
A step towards on-device open-weight AI, though no glasses are announced yet. 👓
PrismML has developed its compact 2B parameter LLMs for Qualcomm Snapdragon-powered smart glasses, shrinking larger models 4x while keeping performance.
A step towards on-device open-weight AI, though no glasses are announced yet. 👓
September 24, 2026 at 7:58 PM
PrismML's tiny LLMs hit smart glasses
PrismML has developed its compact 2B parameter LLMs for Qualcomm Snapdragon-powered smart glasses, shrinking larger models 4x while keeping performance.
A step towards on-device open-weight AI, though no glasses are announced yet. 👓
PrismML has developed its compact 2B parameter LLMs for Qualcomm Snapdragon-powered smart glasses, shrinking larger models 4x while keeping performance.
A step towards on-device open-weight AI, though no glasses are announced yet. 👓
iPhone 17 Proで270億パラメータのAI「Qwen3.6-27B」が動作、AppleはPrismMLと圧縮技術の活用を協議か
https://gigazine.net/news/20260710-prismml-iphone-apple-qwen/
https://gigazine.net/news/20260710-prismml-iphone-apple-qwen/
iPhone 17 Proで270億パラメータのAI「Qwen3.6-27B」が動作、AppleはPrismMLと圧縮技術の活用を協議か
AI開発企業のPrismMLが270億個のパラメータを持つ大規模言語モデル「Qwen3.6-27B」を圧縮し、iPhone 17 Pro単体で動作させることに成功したと主張しています。また、AppleがPrismMLと会合を開き、PrismMLの技術をiPhoneで利用する方法について話し合ったとThe Informationが報じています。
gigazine.net
July 10, 2026 at 6:10 AM
iPhone 17 Proで270億パラメータのAI「Qwen3.6-27B」が動作、AppleはPrismMLと圧縮技術の活用を協議か
https://gigazine.net/news/20260710-prismml-iphone-apple-qwen/
https://gigazine.net/news/20260710-prismml-iphone-apple-qwen/
PrismML Snapdragon için 1-bit Bonsai LLM modelini tanıttı
PrismML, Qualcomm’un Snapdragon platformu için optimize edilmiş 1-bit Bonsai dil modelini duyurdu. Model akıllı gözlüklerde yerel olarak çalışabiliyor.
haber.com
PrismML, Qualcomm’un Snapdragon platformu için optimize edilmiş 1-bit Bonsai dil modelini duyurdu. Model akıllı gözlüklerde yerel olarak çalışabiliyor.
haber.com
September 25, 2026 at 6:53 PM
PrismML Snapdragon için 1-bit Bonsai LLM modelini tanıttı
PrismML, Qualcomm’un Snapdragon platformu için optimize edilmiş 1-bit Bonsai dil modelini duyurdu. Model akıllı gözlüklerde yerel olarak çalışabiliyor.
haber.com
PrismML, Qualcomm’un Snapdragon platformu için optimize edilmiş 1-bit Bonsai dil modelini duyurdu. Model akıllı gözlüklerde yerel olarak çalışabiliyor.
haber.com
PrismML releases Bonsai 27B, claiming first major AI model of its size fit for iPhone
PrismML releases Bonsai 27B, claiming first major AI model of its size fit for iPhone
PrismML is back in the news today after the AI startup’s CEO told CNBC that Apple is looking at the company’s technology. The company has also released its Bonsai 27B model that it says runs on iPhone, iPad, and Mac.
9to5mac.com
July 14, 2026 at 6:12 PM
PrismML releases Bonsai 27B, claiming first major AI model of its size fit for iPhone