Mr. Brooks Whale (@Nely_Brooks) header
Mr. Brooks Whale (@Nely_Brooks) profile picture

🤖 AI Tools | News | Innovation 🚀 Exploring the future of AI 📩 DM for Collaborations & Promotions [email protected]

New York, USA
Joined June 2017
Photo by Mr. Brooks Whale (@Nely_Brooks)Photo by Mr. Brooks Whale (@Nely_Brooks)Photo by Mr. Brooks Whale (@Nely_Brooks)Photo by Mr. Brooks Whale (@Nely_Brooks)Photo by Mr. Brooks Whale (@Nely_Brooks)Photo by Mr. Brooks Whale (@Nely_Brooks)Photo by Mr. Brooks Whale (@Nely_Brooks)Photo by Mr. Brooks Whale (@Nely_Brooks)Photo by Mr. Brooks Whale (@Nely_Brooks)Photo by Mr. Brooks Whale (@Nely_Brooks)
This is an interesting combination of capabilities. LingSante-3.1-air-VL can interpret medical images, generate diagnostic-style reports, search medical literature, and evaluate evidence. It also demonstrates strong general-purpose multimodal performance. Specialized AI models are getting seriously capable.
Meet LingSante-3.1-air-VL, an open-source multimodal medical model with 124B total and 5.5B active parameters. It combines medical image understanding, report generation and medical reasoning with multimodal search and tool use. In our evaluation, it leads the compared models on AFUMED-Derm, AFUMED-Report, MIMIC-CXR and medical VQA (averaged across benchmarks), ahead of GPT-6.1 Sol, Gemini 3.8 Flash on all four. It also demonstrates strong general-purpose capabilities, scoring higher than GPT-6.1 Sol on BrowseComp-VL and Gemini 3.8 Flash on AndroidWorld, alongside competitive performance on MMSearch. Explore LingSante-3.1-air-VL and build with it.
56
Mr. Brooks Whale retweeted
Meet LingSante-3.1-air-VL, an open-source multimodal medical model with 124B total and 5.5B active parameters. It combines medical image understanding, report generation and medical reasoning with multimodal search and tool use. In our evaluation, it leads the compared models on AFUMED-Derm, AFUMED-Report, MIMIC-CXR and medical VQA (averaged across benchmarks), ahead of GPT-6.1 Sol, Gemini 3.8 Flash on all four. It also demonstrates strong general-purpose capabilities, scoring higher than GPT-6.1 Sol on BrowseComp-VL and Gemini 3.8 Flash on AndroidWorld, alongside competitive performance on MMSearch. Explore LingSante-3.1-air-VL and build with it.
29
129
173
420
149,228
Mr. Brooks Whale retweeted
Another major milestone for open-weight AI. LingSante-3.1-flash brings 560B parameters and specialized medical capabilities into one model. From complex reasoning to long-horizon healthcare tasks, the possibilities for developers are getting much bigger.
Today, we release LingSante-3.1-flash, an open-weight MoE model with 560B total parameters and 25B activations per token. It unifies diverse medical capabilities while retaining general capabilities, including complex medical reasoning, long-horizon workflow execution, and professional behavior alignment. LingSante-3.1-flash moves beyond medical QA toward real-world healthcare problem. The model leads on HealthBench Professional (65.72), HLE (text-only, with tools: 72.22) and BrowseComp (with context: 94.10), outperforming GPT-6 Astra, Claude Opus 5.5 and Gemini 4 Argon on all three. LingSante-3.1-flash also performs well on professional long-horizon medical workflows, such as PhysicianBench (57) and EHR-Complex (74.40). Together, these results establish LingSante-3.1-flash as the leading open-weight models for healthcare. The weights are available. Explore LingSante and build with it. More below ↓
7
14
14
181
Mr. Brooks Whale retweeted
A 124B parameter medical AI model with only 5.5B active parameters. That's already interesting. But LingSante-3.1-air-VL also brings image understanding, medical reasoning, evidence search, and tool interaction together. Open-source multimodal AI keeps raising the bar.
Meet LingSante-3.1-air-VL, an open-source multimodal medical model with 124B total and 5.5B active parameters. It combines medical image understanding, report generation and medical reasoning with multimodal search and tool use. In our evaluation, it leads the compared models on AFUMED-Derm, AFUMED-Report, MIMIC-CXR and medical VQA (averaged across benchmarks), ahead of GPT-6.1 Sol, Gemini 3.8 Flash on all four. It also demonstrates strong general-purpose capabilities, scoring higher than GPT-6.1 Sol on BrowseComp-VL and Gemini 3.8 Flash on AndroidWorld, alongside competitive performance on MMSearch. Explore LingSante-3.1-air-VL and build with it.
5
9
9
237
Mr. Brooks Whale retweeted
Imagine an AI that can analyze sleep data, search medical literature, and reason through complex healthcare scenarios. That's the direction LingSante-3.1-flash is taking. 560B total parameters and open weights make this launch worth watching.
Today, we release LingSante-3.1-flash, an open-weight MoE model with 560B total parameters and 25B activations per token. It unifies diverse medical capabilities while retaining general capabilities, including complex medical reasoning, long-horizon workflow execution, and professional behavior alignment. LingSante-3.1-flash moves beyond medical QA toward real-world healthcare problem. The model leads on HealthBench Professional (65.72), HLE (text-only, with tools: 72.22) and BrowseComp (with context: 94.10), outperforming GPT-6 Astra, Claude Opus 5.5 and Gemini 4 Argon on all three. LingSante-3.1-flash also performs well on professional long-horizon medical workflows, such as PhysicianBench (57) and EHR-Complex (74.40). Together, these results establish LingSante-3.1-flash as the leading open-weight models for healthcare. The weights are available. Explore LingSante and build with it. More below ↓
8
16
16
127
Mr. Brooks Whale retweeted
Open-source medical AI just took another big step. LingSante-3.1-flash can handle everything from medical literature searches to complex clinical reasoning. What makes this exciting is the combination of powerful capabilities and open weights that developers can actually explore.
Today, we release LingSante-3.1-flash, an open-weight MoE model with 560B total parameters and 25B activations per token. It unifies diverse medical capabilities while retaining general capabilities, including complex medical reasoning, long-horizon workflow execution, and professional behavior alignment. LingSante-3.1-flash moves beyond medical QA toward real-world healthcare problem. The model leads on HealthBench Professional (65.72), HLE (text-only, with tools: 72.22) and BrowseComp (with context: 94.10), outperforming GPT-6 Astra, Claude Opus 5.5 and Gemini 4 Argon on all three. LingSante-3.1-flash also performs well on professional long-horizon medical workflows, such as PhysicianBench (57) and EHR-Complex (74.40). Together, these results establish LingSante-3.1-flash as the leading open-weight models for healthcare. The weights are available. Explore LingSante and build with it. More below ↓
7
16
16
244
Mr. Brooks Whale retweeted
🚨号党狂喜!多账号AI管理神器开源了 Antigravity-Manager 核心功能速览 ✅ 账号管理:多账号配额实时监控、一键切换、403异常自动检测 ✅ 协议转换:OpenAI / Anthropic / Gemini 三协议兼容,主流工具直连 ✅ 智能调度:限流自动轮换、自定义模型路由、后台任务降级省额 ✅ 多端部署:桌面客户端、Docker 服务端、Termux 移动端 ✅ 安全可控:完全开源,本地部署,账号密钥不外泄 项目地址:github.com/lbjlaq/Antigravit…
🚨憋了三年的王炸!摩尔线程庐山显卡官宣:48G显存,性能接近RTX 5080 摩尔线程MTT S80产品经理AIR_FORCE B站动态实锤,时隔三年回归“务正业”,新一代庐山显卡即将发布: - 48GB超大显存,直接刷新消费级显卡历史最大显存纪录 ​ - 性能足够惊艳,整体表现对标英伟达RTX 5080级别 ​ - 主打大显存刚需市场,错过RTX 5090、有本地大模型/科研需求的用户可以重点蹲守 国产显卡首次摸到次旗舰性能门槛
7
1
1
18
1,091
Mr. Brooks Whale retweeted
细思极恐的一幕:有人让 Gemini 4 Neon 自己挑一首歌唱,并用动态 ASCII 字符画演出它思考时的“内心戏”,全程不准用任何音频素材和 TTS 语音包。 结果它纯靠写代码给自己搓了副声带,在几百万首歌里偏偏自己挑了《Daisy Bell》——这正是科幻片《2001太空漫游》里失控 AI HAL 9000 被拔电源前唱的最后一首歌,也是 1961 年人类历史上计算机唱出的第一首歌🔥 看着那张由代码字符拼出来的脸一边唱情歌一边闪过内心戏,评论区全看进赛博恐怖谷了。
7
2
13
1,758
Mr. Brooks Whale retweeted
This is WAN 3.0 doing a full 30 second scene from one prompt. Three shots, one cook, same stall and rain all the way through, with sound built in. Made with WAN 3.0 on PixelDojo. Try it here:pixeldojo.ai/wan-3-0-video?a…
9
21
24
435
Mr. Brooks Whale retweeted
Creative challenge: make a regular business feel like it has its own GTA 6 introduction. Not a gameplay clip. Not a standard product montage. Something in between. This is what I came up with using GPT-6.2 Astra + CapCut. capcut.com/tools/capcut-x-co… #GPT62Astra #GTA6
3
15
16
439
Mr. Brooks Whale retweeted
Getting "WASTED" in GTA back in the day: - Get hit by a truck. - Wake up outside the hospital 5 seconds later. - Pay $100. - Full health, zero debt. Done. Getting "WASTED" in GTA in 2026: ↓↓
7
11
13
1,376
Mr. Brooks Whale retweeted
This is how dentists restores a deep decayed tooth using modern composite filling technique for a natural looking results
4
17
21
988
Mr. Brooks Whale retweeted
Seedance 2.5 is turning imagination into cinematic reality. 🎬✨ From breathtaking visuals to fluid motion and storytelling that feels alive, AI video is pushing creative boundaries like never before. The only limit now is your imagination. 🚀
12
45
57
648
Mr. Brooks Whale retweeted
Hello, Stop Drama, Build Reputations Family! Happy to join this amazing community! Let's connect, follow each other, spread positivity, and grow through genuine support and mutual respect! more trust, stronger connections! Much love from @TGWorldPlus! nitter.cf/i/spaces/1RekXAEBzVRvQ…

1
4
9
113
Mr. Brooks Whale retweeted
That idea in your head? It doesn't have to stay there. Give it a prompt, explore a few styles, and see what you can create. Buy or upgrade to CapCut Web Ultra between Nov 20–26 and get 7 days of unlimited image generation on select models. More room to experiment means more opportunities to turn a random thought into something worth sharing. Let's make those ideas happen! link:capcut.com/my-edit?utm_mediu… #CapCutWeb #CapCutWebDesignStudio #CapCutWebVideoEditor #aiimage
🤝 Paid partnership
3
12
14
159
Mr. Brooks Whale retweeted
This is more than another large language model announcement. LingSante-3.1-flash is built around real healthcare challenges, including medical research, data analysis, and complex reasoning. Open-weight models are becoming increasingly capable of tackling specialized work.
Today, we release LingSante-3.1-flash, an open-weight MoE model with 560B total parameters and 25B activations per token. It unifies diverse medical capabilities while retaining general capabilities, including complex medical reasoning, long-horizon workflow execution, and professional behavior alignment. LingSante-3.1-flash moves beyond medical QA toward real-world healthcare problem. The model leads on HealthBench Professional (65.72), HLE (text-only, with tools: 72.22) and BrowseComp (with context: 94.10), outperforming GPT-6 Astra, Claude Opus 5.5 and Gemini 4 Argon on all three. LingSante-3.1-flash also performs well on professional long-horizon medical workflows, such as PhysicianBench (57) and EHR-Complex (74.40). Together, these results establish LingSante-3.1-flash as the leading open-weight models for healthcare. The weights are available. Explore LingSante and build with it. More below ↓
9
17
21
435
Mr. Brooks Whale retweeted
🚨 BREAKING: OpenAI fires a software engineer earning over $300,000 for "cheating"! A leaked official email revealed that the company has terminated one of its senior engineers after system records showed he had no real understanding of the source code. The irony? 100% of the code he submitted was generated by Claude, he used ChatGPT daily to explain Claude's code, and he used Gemini to fix all the bugs. And during a critical production crisis yesterday, all he did was type the following to ChatGPT: "Please fix this, they're watching me right now". His only defense: "I was just a human interface between the three models, and the system was working perfectly until today".
2
18
22
570
Mr. Brooks Whale retweeted
The real value of medical AI isn't just answering faster. It's understanding complex cases and finding the right evidence. LingSante-3.1-flash is pushing toward that vision with advanced reasoning and professional healthcare workflow capabilities.
Today, we release LingSante-3.1-flash, an open-weight MoE model with 560B total parameters and 25B activations per token. It unifies diverse medical capabilities while retaining general capabilities, including complex medical reasoning, long-horizon workflow execution, and professional behavior alignment. LingSante-3.1-flash moves beyond medical QA toward real-world healthcare problem. The model leads on HealthBench Professional (65.72), HLE (text-only, with tools: 72.22) and BrowseComp (with context: 94.10), outperforming GPT-6 Astra, Claude Opus 5.5 and Gemini 4 Argon on all three. LingSante-3.1-flash also performs well on professional long-horizon medical workflows, such as PhysicianBench (57) and EHR-Complex (74.40). Together, these results establish LingSante-3.1-flash as the leading open-weight models for healthcare. The weights are available. Explore LingSante and build with it. More below ↓
14
21
21
312
Mr. Brooks Whale retweeted
What stands out about LingSante-3.1-flash is its focus on real-world healthcare challenges. From sleep analysis to medical research and clinical reasoning, it's designed for much more than answering basic questions. Open-source medical AI keeps getting stronger.
Today, we release LingSante-3.1-flash, an open-weight MoE model with 560B total parameters and 25B activations per token. It unifies diverse medical capabilities while retaining general capabilities, including complex medical reasoning, long-horizon workflow execution, and professional behavior alignment. LingSante-3.1-flash moves beyond medical QA toward real-world healthcare problem. The model leads on HealthBench Professional (65.72), HLE (text-only, with tools: 72.22) and BrowseComp (with context: 94.10), outperforming GPT-6 Astra, Claude Opus 5.5 and Gemini 4 Argon on all three. LingSante-3.1-flash also performs well on professional long-horizon medical workflows, such as PhysicianBench (57) and EHR-Complex (74.40). Together, these results establish LingSante-3.1-flash as the leading open-weight models for healthcare. The weights are available. Explore LingSante and build with it. More below ↓
9
15
16
347
Mr. Brooks Whale retweeted
POV: You opened GTA 6 to play, but somehow ended up shopping. 💀 This business really got the full game-world introduction treatment. The character reveal, the product selection screen... even the ad feels like a little side quest. capcut.com/tools/capcut-x-co… #GTA6 #CapCutDidThat
🤝 Paid partnership
5
21
22
486