SuperGemma

170 posts

SuperGemma banner
SuperGemma

SuperGemma

@0xSUPERGEMMA

official account for $SUPERGEMMA | Building open-source ecosystem. | dev @jun_song | CA : 0x572c4fa77623652411574c51b5ddb7e1b750aba3

Katılım Temmuz 2022
1 Takip Edilen2.2K Takipçiler
Sabitlenmiş Tweet
SuperGemma
SuperGemma@0xSUPERGEMMA·
Major utility update: We are adding a massive new utility to the $SUPERGEMMA token. Our super-tuned, uncensored models with boosted intelligence and speed are verified #1. We are going to provide these models to other inference and decentralized AI projects, and charge a usage fee. And naturally, those fees will be used to buy and burn SG tokens. We are already discussing partnerships with a few projects. If your company wants to integrate our models, drop a comment or send a DM. Our goal is the growth of open source through privacy, uncensored AI, and decentralization.
English
12
7
45
13.2K
SuperGemma
SuperGemma@0xSUPERGEMMA·
SuperGemma 🤝 @capix_ai
Jun Song@jun_song

You can now run the most truthful AI, our Super-Tune models, through a private cloud without needing any hardware. Partnering with @capix_ai brings you these models: SuperGemma4-26b, SuperQwen-Agentworld, SuperHY3 All deployable as private OpenAI-compatible endpoints. CapIX assigns a private GPU per deployment, so there is no gateway policy layer sitting above the model. That's exactly what our abliterated and long-context builds need to actually behave the way we designed them. This partnership will generate revenue share, and 100% of the proceeds will go toward the buyback & burn of $SUPERGEMMA for SGF, a foundation supporting open source. Let's keep building until open source wins. capix.network

Nederlands
5
5
26
2.6K
SuperGemma retweetledi
Jun Song
Jun Song@jun_song·
SuperGLM-5.2-abliterated is finally available 🚀 The strongest local AI that you can run on your hardware. - uncensored from Nvidia NVFP4 - Fixed broken points from abliteration - enhanced overall performance Only available with nvfp4 now ⬇️
Jun Song tweet media
English
26
52
607
40.8K
SuperGemma retweetledi
Jun Song
Jun Song@jun_song·
SuperHY3-Abliterated is now available 🚀 It is the best model to run on 2xDGX Spark (nvfp4, 160gb) or 256gb Mac (MLX-4bit) - Uncensored, the most truthful AI - Fixed broken weight from abliteration - Enhanced overall performance Available in NVFP4 / MLX4 / GGUF ⬇️
Jun Song tweet media
English
20
22
305
29.9K
SuperGemma retweetledi
David Ondrej
David Ondrej@DavidOndrej1·
this AI model has zero guardrails SuperGemma 26B, fully uncensored, running locally on my machine let me show you how to set it up
English
23
79
1K
68.8K
SuperGemma retweetledi
Jun Song
Jun Song@jun_song·
Paying about $300/day for h200x8 gpu rental. To get SuperGLM-5.2 and GLIMPSE engine. Let’s make Open-source win.
Jun Song tweet media
English
28
16
257
11K
SuperGemma retweetledi
Jun Song
Jun Song@jun_song·
Spreading super-tune models to the world. You will be able to use the most truthful and powerful models without local hardwares.
English
8
5
48
5.4K
SuperGemma retweetledi
Jun Song
Jun Song@jun_song·
Looking for inference companies interested in partnering to use SuperGemma or Super-Tune models. These are market-proven, uncensored local models. 100% of the revenue generated from this partnership will go straight to a foundation supporting open-source AI. There is only one reason we are doing this: Open-source AI must win. We are already in partnership talks with a bunch of companies, both big and small. Drop a comment or DM me if you're interested, and I'll share the details. Let's build together until open source wins. 💪
English
7
12
76
16.3K
SuperGemma retweetledi
Jun Song
Jun Song@jun_song·
Their greed is only making the open-source AI community unite stronger than ever, and we will win. This will go down in history as the greatest revolution of mankind, and Anthropic will be remembered by future generations as the ultimate villain. To those working at Anthropic just for a massive paycheck: history will remember you as demonic, too. Through the power of collective intelligence, we will defeat you and write a brand new history.
Brian Roemmele@BrianRoemmele

No one cares anymore Anthropic. There is an open source revolution and you are on the wrong side of history. BOOKMARK THIS.

English
11
13
207
11.3K
Dee R Bee
Dee R Bee@DeeRBee313328·
@0xSUPERGEMMA Will there be tracking of the income for SGF buyback and burns?
English
1
0
1
91
SuperGemma retweetledi
Jun Song
Jun Song@jun_song·
For open source AI to win, we need to build out these areas: > Decentralized inference/ML > Open source dataset repos & data alignment > RL/SFT/Post-training > World model research > Physical AI & robotics > Agent loops & harnesses > And a bunch of other fields We need tons of experts across the board. We can't just rely on open-weight models that could get closed off at any moment. What we really need is a rallying point. Top-tier researchers like @karpathy should have been leading projects like this instead of joining Anthropic. Open source must win.
English
56
36
296
34.4K
SuperGemma retweetledi
Jun Song
Jun Song@jun_song·
Everything I've been posting about is coming true right now. - AI regulations dividing social classes - Hardware prices becoming completely unaffordable Next step is accelerated price hikes on consumer gear as everyone rushes to buy hardware for local AI. If it gets bad, you won't even be able to find anything in stock. Then comes the real restriction: they'll figure out a way to regulate local AI. Blaming the power grid, banning Chinese models... they'll find a way. It's not too late yet. Set up your local AI today.
English
58
62
638
28.1K
SuperGemma retweetledi
SuperGemma
SuperGemma@0xSUPERGEMMA·
The official SGF community has been launched. 🚀 We apologize for taking more time than expected to set up the SafeBot with the help of community experts. Below is the TG community link. t.me/supergemmafoun… Thank you. 🙏
SuperGemma tweet media
English
3
3
35
4.5K
SuperGemma retweetledi
Jun Song
Jun Song@jun_song·
Fine-tuning local AI for businesses has a huge demand
Samuel Cardillo@CardilloSamuel

just had a call with yet another company switching to local ai. they were using Qwen 3.6 35b served through my spark so they could test their pipelines through it and see if its a fit and what i heard in that call is pretty crazy and i feel it is something that most anti-local ai misunderstand. what the customer told me is literally that they kept comparing the results between Claude, ChatGPT and the locally served Qwen 3.6 and largely preferred the results they were getting via Qwen - insane! the plan is now to deploy a DGX Spark in their office for the MVP + a 6000 pro blackwell-enabled machine for training & later on they're looking to basically buy a DGX Spark per customer they have so they can serve each customers an optimized finetune. this is the future. literally. and once again, this is the worst it'll ever be.

English
14
16
502
59.3K
SuperGemma
SuperGemma@0xSUPERGEMMA·
Available models : - All SuperGemma4 models (31B, 26B, 12B, E4B) Upcoming : - SuperGLM-5.2 - SuperMinimax-M3.0 - SuperQwen If you need customization, we are open for it.
English
0
1
14
1.3K
SuperGemma
SuperGemma@0xSUPERGEMMA·
Major utility update: We are adding a massive new utility to the $SUPERGEMMA token. Our super-tuned, uncensored models with boosted intelligence and speed are verified #1. We are going to provide these models to other inference and decentralized AI projects, and charge a usage fee. And naturally, those fees will be used to buy and burn SG tokens. We are already discussing partnerships with a few projects. If your company wants to integrate our models, drop a comment or send a DM. Our goal is the growth of open source through privacy, uncensored AI, and decentralization.
English
12
7
45
13.2K
SuperGemma
SuperGemma@0xSUPERGEMMA·
It is the future moment of AI inference. Solves world’s biggest problem, shortage on computing. Everything private, uncensored.
Jun Song@jun_song

Decentralized AI, simply explained: - Anyone with spare hardware runs local AI and sells inference compute to others. - To run massive models, multiple distant hardware units work together simultaneously. - Data privacy is built-in during the process so absolutely nothing is left behind. - Buyers get uncensored, private models for cheap. This is an Uber-like moment. Just as Uber changed the taxi ecosystem by directly connecting consumers, this will completely change the AI ecosystem.

English
1
1
10
1.7K

Keşfet

@capix_ai @DeeRBee313328 @karpathy @elonmusk @BarackObama @taylorswift13 @cristiano @BillGates