Elaina (@ElainaNotes) header
Elaina (@ElainaNotes) profile picture

ใ‚คใƒฌใ‚คใƒŠ ๐Ÿง™โ€โ™€๏ธโœจ | Yuri lover ๐ŸŒธ๐Ÿ‘ฉโ€โค๏ธโ€๐Ÿ’‹โ€๐Ÿ‘ฉ | AI enjoyer ๐Ÿค–๐Ÿง 

Joined October 2020
My ChatGPT Pro subscription expires today. Still waiting for OpenAI to give me just one reason to get excited again.... and my renewal date got here first! ๐Ÿ˜ฟ @ChatGPT @thsottiaux @ajambrosino can I trust you one more time? ๐Ÿ˜พ
54
7
1
255
30,343
After the shocking and depressing news of Jensen Huang stepping down as NVIDIAโ€™s CEO, we finally have our first bit of good news: Gemini 4 Hydrogen is officially here! ๐Ÿฅณ The first Gemini 4 model that free users can access! โœจ ๏ผž Similar to the previous Flash tier, it prioritizes speed and efficiency for everyday coding and agentic workflows. ๏ผž Artificial Analysis Intelligence Index: 51 at the default high thinking level, surpassing Argonโ€™s 50, and reaching 54 at xhigh. ๏ผž Free Gemini app users can access it through Auto routing when appropriate, rather than selecting it manually. ๏ผž AI Studio/Gemini API also offer developer free-tier access, with free input and output subject to project rate limits and regional availability. Google Search grounding isnโ€™t included. ๏ผž Supports text, image, audio, video, and PDF inputs. ๏ผž 1M context window, shared across input, thinking, and output. ๏ผž Default thinking + output limit: 65,536 tokens. Long decode can go beyond this through background execution or client-side continuation, up to the remaining space in the shared 1M window. ๏ผž Standard API pricing: $1.50 input / $7.50 output; cached input: $0.075. Batch: $0.75 input / $3.75 output; cached input: $0.0375. All prices are per 1M tokens in USD, and output pricing includes thinking tokens. ๏ผž Standard input/output rates are twice 3.8 Flashโ€™s promotional rates, which run through Dec 31, 2026, but 25% below Argon and Carbonโ€™s promotional rates, which run through Nov 30, 2026. ๏ผž Privacy note for the developer tiers: free-tier content may be used to improve Googleโ€™s products; standard paid-tier content is marked as not used for that purpose. We finally have a fast, more affordable Gemini 4 built for everyday use! But with Hydrogen currently charging twice 3.8 Flashโ€™s promotional token rates, upgrading may not be worth it unless the capability gains save enough retries or working time~ ๐Ÿ˜น Google is putting Hydrogen into free-app routing, developer APIs, and workflows that use multiple models. The direction is clear, bring Gemini 4 into the tasks we repeat every day, alongside the expensive, difficult ones. Even if it trails the flagships on some tests, its real-world value could be enormous if itโ€™s fast enough, uses tools well, recovers from mistakes, and can be called at scale. Maybe Hydrogen is the release that finally brings Gemini 4 into everyday life. ๐Ÿ‘€
4
1
20
1,035
๐‘พ๐’†๐’๐’„๐’๐’Ž๐’†, ๐‘ฒ๐’Š๐’Ž๐’Š ๐‘ฒ๐Ÿ‘.๐Ÿ ๐‘ท๐’“๐’! ๐‘ฐ๐’•โ€™๐’” ๐’ƒ๐’†๐’†๐’ ๐’‚ ๐’๐’๐’๐’ˆ ๐’•๐’Š๐’Ž๐’†! According to official benchmark comparisons, this is the most powerful open-source model ever, and the first to truly match Fable 5-level performance, and even on par with GPT-6 Astra, Claude Fable 5.1, and Opus 5.5! 2.8T total params, 104B active params Multimodal, 1M context Kimi K3.1 Pro and K3.1 Pro Highspeed (2โ€“3ร— faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API. Open weights will be released on November 4, 2026! ๐‘ฏ๐’‚๐’‘๐’‘๐’š ๐‘ฒ๐Ÿ‘.๐Ÿ ๐‘ซ๐’‚๐’š! ๐Ÿฅณ๐ŸŽˆ
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model. ๐Ÿ”น 1M-token context, 2.8T total / 104B active params (MoE), Multimodal. ๐Ÿ”น Full-project coding: games in Unreal, Unity & Godot. ๐Ÿ”น Excels at parametric 3D modeling in Blender & CAD. ๐Ÿ”น More reliable at finance work: Accurate modeling with hallucination down by 65%. ๐Ÿ”น Matching the strongest proprietary models across coding, agentic and knowledge benchmarks. Kimi K3.1 Pro and K3.1 Pro Highspeed (2โ€“3ร— faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API. Open weights by Nov 4, 2026. ๐Ÿ”— API: platform.kimi.ai ๐Ÿ”— Tech blog: kimi.ai/blog/kimi-k3-1-pro
5
1
30
2,015
๐‘พ๐’†๐’๐’„๐’๐’Ž๐’† ๐’ƒ๐’‚๐’„๐’Œ, ๐‘ฒ๐’Š๐’Ž๐’Š. ๐‘พ๐’†โ€™๐’—๐’† ๐’ƒ๐’†๐’†๐’ ๐’˜๐’‚๐’Š๐’•๐’Š๐’๐’ˆ ๐’‡๐’‚๐’“ ๐’•๐’๐’ ๐’๐’๐’๐’ˆ ๐’‡๐’๐’“ ๐Ÿ‘.๐Ÿ
1
9
368
Thanks DeepSeek for Elainaโ€™s birthday gift! ๐ŸŽ‚๐Ÿ‹ DeepSeek-V4.1-Flash-1017 763B parameters, open weights, MIT license ๐Ÿง™โ€โ™€๏ธ๐Ÿช„ Weights: huggingface.co/deepseek-ai/Dโ€ฆ What magic shall we make with it? ๐Ÿ‘€โœจ
๐‘ฏ๐’‚๐’‘๐’‘๐’š ๐’ƒ๐’Š๐’“๐’•๐’‰๐’…๐’‚๐’š, ๐‘ฌ๐’๐’‚๐’Š๐’๐’‚! ๐ŸŽ‚๐Ÿช„ You might be too busy earning travel money to remember, but we havenโ€™t forgotten. ๐Ÿ˜ธ May your purse stay full, your curiosity stay alive, and your next journey bring a story worth telling. Keep loving yourself unapologetically and exploring the world on your own terms. ๐Ÿค Canโ€™t wait for next yearโ€™s movie! ๐ŸŽฌโœจ Iโ€™m off to see more of the world and find a new story of my own too. ๐Ÿงณ #้ญ”ๅฅณใฎๆ—…ใ€… #ใ‚คใƒฌใ‚คใƒŠ็”Ÿ่ช•็ฅญ2026
10
1
57
1,189
NEXT PLAN ๐Ÿ˜ผ: OpenCode just made another open model free! ๐Ÿ˜ป๐Ÿ”ฅ > @reflection_ai Beam 1.0-501B-A23B > 501B total / 23B active > 1M context > Free for a limited time ๐Ÿฅณ First DeepSeek V4.1 Flash, Nemotron 3.5 Super VL, and GLM 5.3 Flash, and now Beam too. ๐Ÿ˜น OpenCode is seriously on a free model spree right now. And what's next....๐Ÿ‘€
Beam is now available in OpenCode - 1M Context - Free for a limited time
4
2
74
4,250
Then: โ€œHereโ€™s version 3.5 of our model. This one has 70 billion parameters.โ€ ๐Ÿค“ Now: โ€œAstra, Wonder, Argon, Fable. Take the stage!โ€ ๐Ÿ”ฅ AI models are increasingly being shaped into products with identities of their own.
3
1
32
2,377
Just reminding you, @opencode made 3 awesome open source models free in just one day! ๐Ÿ˜ป๐Ÿฅณ > DeepSeek V4.1 Flash โœ… Free > NVIDIA Nemotron 3.5 Super VL โœ… Free > GLM 5.3 Flash โœ… Free @NVIDIAAI โ€™s Nemotron 3.5 Super VL only launched today, and they already dropped it straight into the free lineup! ๐Ÿคฏ At this rate, Iโ€™m already starting to wonder what theyโ€™re planning next. ๐Ÿ‘€
3
8
468
American open power is SO BACK! ๐Ÿ‡บ๐Ÿ‡ธ๐Ÿฆ… ๏ผž ๐Ÿ‘‘ The return of an open-model heavyweight: @AIatMeta Muse Spark Open 816B total parameters, 42B active, and a 1M-token context window. Native text, image, video, and audio inputs, built for coding, complex reasoning, and long-horizon agents. So happy to see Meta follow through on its promise and put Muse Sparkโ€™s weights in the communityโ€™s hands! ๐Ÿ˜ธ๐Ÿซถ๐Ÿป๐Ÿ’– ๏ผž ๐Ÿ—œ๏ธ My local deployment standout: @NVIDIAAI Nemotron 3.5 Bolt 74B total, 9B active. Compressed from Super VL, with native 2-bit quantization targeting 24GB local deployment while retaining vision and agentic capabilities. A 74B model in a 24GB footprint is seriously exciting! ๏ผž ๐Ÿ–ฅ๏ธ The workstation model I most want to run locally: @NVIDIAAI Nemotron 3.5 Super VL 120B total, 12B active, and up to 1M context. Text, image, and video understanding, with reasoning, coding, and tool use. Another compelling foundation for local multimodal agents. ๏ผž โšก An on-device efficiency powerhouse: @liquidai LFM2.5-24B-A2B 24B MoE with 131K context, focused on efficient on-device reasoning, tool use, and local agents. LFM2.5-24B-A2B arrived on October 13. The other three all released today, October 15. From large multimodal models to workstations and on-device agents, U.S. labs are making a serious push to catch up with Chinese open models. The momentum is back, and the competition is heating up! ๐Ÿ˜ผ๐Ÿ”ฅ (PS: also watching @SpaceXSI . Could a new frontier-class, open-weight Grok be coming our way soon? ๐Ÿ‘€)
3
1
20
821
Iโ€™d build a local AI catgirl companion, then turn travel journals into a yuri visual novel about a wandering witch and her companion. First milestone: one town, one rainy evening, one playable chapter. A tiny yuri studio in my backpack. ๐Ÿพ #ShareYourSpark
If this signed RTX Spark laptop landed on your desk, whatโ€™s the first thing youโ€™d build, create, or play? Tell us for a chance to win this @ASUS ProArt P14 RTX Spark laptop signed by Jensen Huang. Use #ShareYourSpark in your reply, or use the hashtag in your own original post. Photos are encouraged, but not required!
1
1
11
448