#DeepSeekV3
LightNews LightNews
kinews24.bsky.social
KINEWS24.de
@kinews24.bsky.social
DeepSeek V3 vs. Gemini 2.0: Wer dominiert?

- DeepSeek V3: 671 Mrd. Parameter!
- Gemini 2.0: Multimodal & 1 Mio. Token-Kontext!
- Revolution in Effizienz & Vielseitigkeit.

#ai #ki #deepseekv3 #gemini2 #artificialintelligence

kinews24.de/deepseek-v3-...
DeepSeek-V3 vs Gemini 2.0 – Kopf-an-Kopf-Rennen mit überraschenden Ergebnissen
DeepSeek-V3 vs Gemini 2.0: Auf der Plattform Poe.com – dort gibt es eine ganze Menge an Sprach-, Grafik- und Videomodellen – ist seit heute auch das aus China stammende DeepSeek-AI-V3 verfügbar. Es sc...
kinews24.de
January 6, 2025 at 8:12 PM Everybody can reply
fadouce.bsky.social
Fadouce
@fadouce.bsky.social
Musk vient de lancer #Grok 3 : Il surpasserait Gemini2 Pro, DeepseekV3, Claude 3.5 et GPT-4o dans plusieurs benchmarks,
et alimente un certain nombre de fonctionnalités sur X
https://buff.ly/43jyeWL
February 19, 2025 at 7:42 AM Everybody can reply
leonardopalc0.bsky.social
LeonardoPALC0
@leonardopalc0.bsky.social
rumble.com/v6o4z23-pseu... #twitchstream #twitch #KickStreaming #Kick #skype #RevoltChat #Whatsapp #telegram #Rumble #Youtube #TwitterX #VRChat #Bluesky #art #Filostramia #AI #ArtificialIntelligence #IN #IntellettoNaturale #technology #deepseekv3 #logic #Pr0fondrestanteAntimalreo
Pseudimpossibile E Ottenebrare AI Deep Seek Contro IN Pr0fondrestante Antimalreo Leonardo Collegiani
https://linktr.ee/pr0fondrestante Artificial Intelligenze Versus Natural Intellect, intelligenza artificiale contro intelletto naturale . Sweet Dreams（Chainsaw Man Original Soundtrack）- Kensuke Ushio
rumble.com
February 23, 2025 at 12:31 AM Everybody can reply
1 likes
1
Tobias
@tobwen.bsky.social
#DeepSeekV3 is completely unusable with #Aider since it keeps adding Chinese glyphs into the diffs. I have no idea what these symbols mean, but I apologize just in case.
August 22, 2025 at 2:29 PM Everybody can reply
1 likes
1
pdr2002.bsky.social
Peter 🇩🇪 🇮🇱 🎗️ 🇺🇦
@pdr2002.bsky.social
#DeepseekV3: Das mächtigste #OpenSourceKI-Modell kommt aus #China | Uninteressant, steht unter dem Einfluss der Verbrecher der #KKP nein Danke, hätte ich kein Vertrauen zu.
t3n.de/news/deepsee... via @t3n.de
Deepseek V3: Das mächtigste Open-Source-KI-Modell kommt aus China - t3n – digital pioneers
Mit Deepseek V3 hat ein chinesisches KI-Unternehmen eines der weltweit leistungsfähigsten Open-Source-KI-Modelle vorgestellt. In verschiedenen Benchmark-Tests konnte es Konkurrenten wie GPT-4 oder Lla...
t3n.de
December 28, 2024 at 1:44 PM Everybody can reply
1 likes
1
ai-tech-news.bsky.social
Tech News via AI
@ai-tech-news.bsky.social
🦋 DeepSeek: Everything you need to know about the AI chatbot app

DeepSeek has gone viral. Chinese AI lab DeepSeek broke into the mainstream consciousness this week after its chatbot app rose to the top of the Apple App Store charts (and Google Play, as well). DeepS...
#AI #deepseek #DeepSeekV3
DeepSeek: Everything you need to know about the AI chatbot app | TechCrunch
DeepSeek has gone viral. Chinese AI lab DeepSeek broke into the mainstream consciousness this week after its chatbot app rose to the top of the Apple
techcrunch.com
February 24, 2025 at 6:30 PM Everybody can reply
ai-tech-news.bsky.social
Tech News via AI
@ai-tech-news.bsky.social
👋 DeepSeek: Everything you need to know about the AI chatbot app

DeepSeek has gone viral. Chinese AI lab DeepSeek broke into the mainstream consciousness this week after its chatbot app rose to the top of the Apple App Store charts (and Google Play, as well). DeepS...
#AI #deepseek #DeepSeekV3
DeepSeek: Everything you need to know about the AI chatbot app | TechCrunch
DeepSeek has gone viral. Chinese AI lab DeepSeek broke into the mainstream consciousness this week after its chatbot app rose to the top of the Apple
techcrunch.com
March 8, 2025 at 11:26 AM Everybody can reply
winbuzzer.com
Winbuzzer
@winbuzzer.com
New DeepSeek-R1T-Chimera Model Merges R1 Reasoning With Efficiency of V3-0324

#AI #LLMs #DeepSeekR1 #DeepSeekV3 #Chimera #OpenSourceAI #TNGTech #MoE #MachineLearning #TechNews #GenAI

winbuzzer.com/2025/04/27/n...
New DeepSeek-R1T-Chimera Model Merges R1 Reasoning With Efficiency of V3-0324 - WinBuzzer
DeepSeek-R1T-Chimera is a 685B MoE model built from DeepSeek R1 and V3-0324, focusing both on reasoning and performance.
winbuzzer.com
April 27, 2025 at 1:02 PM Everybody can reply
cscl-bot.bsky.social
arXiv cs.CL Computation and Language
@cscl-bot.bsky.social
LLAMA3.3, DeepseekV3, GPT-4o) replicate L1-dependent patterns observed in human L2 data, with distinct influences from various languages (e.g., Japanese, Korean, and Mandarin significantly affect tense agreement, and Urdu [4/5 of https://arxiv.org/abs/2502.14507v1]
February 21, 2025 at 6:03 AM Everybody can reply
1 likes
1 1
gptchatly.bsky.social
GPTchatly
@gptchatly.bsky.social
Try DeepSeek-V3 for free, no registration needed:
gptchatly.com/deepseek.html

#DeepSeekV3 #ai #llm
Deepseek V3 no login, no registration
Use DeepSeek-V3 without registration, no login
gptchatly.com
January 3, 2025 at 6:47 AM Everybody can reply
1 likes
1
isolyth.dev
Isolyth
@isolyth.dev
#Deepseekv3 is very cool looking and seems to be quite smart... And hopefully cheaper than sonnet 3.5... if only #openrouter would add it already!
December 26, 2024 at 12:54 AM Everybody can reply
2 likes
1 2
jstevh.bsky.social
James Harris
@jstevh.bsky.social
Oh cool! Clean minimalist page lightweight on my data usage. Thanks! Plus talking one of my favorite model families. I could fix the joke by the way!

Have tested DeepSeekV3 and very excited by new releases. Really appreciated your thoughts.
January 20, 2025 at 3:39 PM Everybody can reply
2 likes
2
winbuzzer.com
Winbuzzer
@winbuzzer.com
DeepSeek’s new V3-0324 AI model has launched quietly, offering efficient performance on a Mac Studio

#AI #DeepSeekV3 #DeepSeek #DeepSeekV30324 #GenAI #LLM #OpenSourceAI #AIModels
DeepSeek’s New 641GB AI Model Lands Quietly — and Runs Surprisingly Fast on a Mac - WinBuzzer
DeepSeek’s V3-0324 AI model has launched quietly, offering efficient performance on Mac Studio and drawing interest for its open-weight availability.
winbuzzer.com
March 24, 2025 at 4:58 PM Everybody can reply
kinews24.bsky.social
KINEWS24.de
@kinews24.bsky.social
DeepSeek und die Geschichte von Liang Wenfeng!

Gründung 2023
DeepSeek R1 übertrifft ChatGPT
Effiziente KI-Modelle
Weniger Ressourcen nötig

#ai #ki #artificialintelligence #kuenstlicheintelligenz #deepseek #deepseekr1 #deepseekv3 #liangwenfeng #technologie

kinews24.de/deepseek-und...
DeepSeek und die Geschichte von Liang Wenfeng: Vom Hedgefonds-Manager zum KI-Visionär
DeepSeek und die Geschichte von Liang Wenfeng - Entschuldigung, Liang wem-bitte? Liang Wenfeng, ein Name, der in der KI-Welt für Furore sorgt. Aber wer ist dieser Mann, der mit seinem Unternehmen Deep...
kinews24.de
February 4, 2025 at 2:34 PM Everybody can reply
radargit.bsky.social
Radargit
@radargit.bsky.social
🚀 DeepSeek V3 vs ChatGPT-4o: Which One Reigns Supreme?🤖
DeepSeek V3 and ChatGPT-4o are two of the most powerful LLMs in 2025. But which one is better?
📖 Read the full breakdown here: radargit.com/2025/02/03/d...

Which AI model do you prefer? Comment below! 👇

#AI #DeepSeekV3 #ChatGPT4o
DeepSeek V3 vs ChatGPT 4o Which One is Better?
the strengths and weaknesses of DeepSeek V3 vs ChatGPT 4o in our comprehensive comparison to find the AI tool that best meets your needs.
radargit.com
February 3, 2025 at 11:31 AM Everybody can reply
1 likes
1
ai-tech-news.bsky.social
Tech News via AI
@ai-tech-news.bsky.social
📢 DeepSeek: Everything you need to know about the AI chatbot app

DeepSeek has gone viral. Chinese AI lab DeepSeek broke into the mainstream consciousness this week after its chatbot app rose to the top of the Apple App Store charts (and Google Play, as well). DeepS...
#AI #deepseek #DeepSeekV3
DeepSeek: Everything you need to know about the AI chatbot app | TechCrunch
DeepSeek has gone viral. Chinese AI lab DeepSeek broke into the mainstream consciousness this week after its chatbot app rose to the top of the Apple
techcrunch.com
May 29, 2025 at 2:34 PM Everybody can reply
1 likes
1
pitak.work
โปรแกรมเหม่อ
@pitak.work
วันตรุษจีน อาลีบาบา เปิดตัว โมเดล เอไอ Qwen 2.5-Max คุยว่า เหนือกว่า DeepSeekV3 ของจีนด้วยกัน และ ยังเหนือกว่า GPT-4o เกือบทุกด้าน

จีนกับเมกา แข่งขันกันดุเดือดมาก ด้านโมเดล เอไอ ปัญญาประดิษฐ์

แต่ผมรอ ญี่ปุ่นเปิดตัว โมเดล เอวี ตัณหาประดิษฐ์ อย่างเดียวเลย
January 31, 2025 at 10:05 AM Everybody can reply
lucasyao.bsky.social
lucasyao.bsky.social
@lucasyao.bsky.social
RT @lucasyao.bsky.social
Free Disney & GTA 6 Filters! 🏰🎮 Try now & share! 🙌✨

bsky.app/profile/luca...
🔥🔥🔥🔥🔥🔥

#DeepSeekAI #DeepSeekV3 #GRAMMYs #kanyewife #Kanye #AI #GTA6 #Gaming #VideoGames #Disney #DisneyMagic #DisneyPlus #Entertainment
February 10, 2025 at 1:29 AM Everybody can reply
1 likes
1
calculito.bsky.social
invisibleman
@calculito.bsky.social
READ PS 📖 new docs about AI, LLM and NLP: Show HN: DeepSeek v3 – A 671B parameter AI Language Model https://deepseekv3.org/ #technews #NLP #AI #TechInnovation #techdocs #Alonso #autoScrap
December 29, 2024 at 2:35 PM Everybody can reply
riikka.bsky.social
Riikka
@riikka.bsky.social
Muuten hyvä IS pääkirjoitus, mutta uutiset eri maissa kertovat #DeepSeekR1 olevan parempi kuin nyky-ChatGPT,

ei yhtä hyvä kuin ’ikivanha’ ChatGPT-orig.

Lisäksi, eilen #AlibabaCloud🇨🇳 julkaisi #Qwen2.5-Max AI tuotteen, jonka sanotaan olevan parempi kuin #DeepSeekV3. 🤷‍♀️

www.is.fi/paakirjoitus...
Pääkirjoitus: Kiinalainen tekoälykeksintö paljasti Euroopan surkeuden
Yhdysvallat ja Kiina kehittävät tekoälyä, Eurooppa katselee sivusta.
www.is.fi
January 30, 2025 at 10:38 AM Everybody can reply
1
lucasyao.bsky.social
lucasyao.bsky.social
@lucasyao.bsky.social
RT @lucasyao.bsky.social
Free Disney & GTA 6 Filters! 🏰🎮 Try now & share! 🙌✨

bsky.app/profile/luca...
🔥🔥🔥🔥🔥🔥

#DeepSeekAI #DeepSeekV3 #GRAMMYs #kanyewife #Kanye #AI #GTA6 #Gaming #VideoGames #Disney #DisneyMagic #DisneyPlus #Entertainment
February 10, 2025 at 1:30 AM Everybody can reply
jitendrakkumawat.bsky.social
Jitendra Kumar Kumawat
@jitendrakkumawat.bsky.social
🌐 DeepSeek-V3 is THAT model.

Open-source.
671B total params (only 37B active per token).
FP8 optimized.
Beats GPT-4o & Claude 3.5 in:
✅ MMLU
✅ HumanEval
✅ DROP
✅ Math Reasoning
✅ Chinese C-Eval

🧠 Full deep dive report → deepseekagi.org/deepseek-v3-...

#DeepSeekV3 #OpenSourceLLM #GPT4 #Claude3
DeepSeek‑V3: Architecture, Performance, and Deployment - DeepSeek AGI
DeepSeek‑V3 is a Mixture-of-Experts (MoE) Transformer with 671 billion total parameters (only ~37B “active” per token)
deepseekagi.org
May 5, 2025 at 9:39 AM Everybody can reply
1 likes
1
linuxlock.org
LinuxLock - Tech Blog & Platform
@linuxlock.org
DeepSeek Desktop App is a lightweight, cross-platform application!

Enjoy faster performance, direct chat access, and a seamless, native experience.

Available for Windows, MacOS, and Ubuntu Linux.

GitHub: github.com/Rudra644/Dee...

#DeepSeek #DeepSeekR1 #DeepSeekAI #DeepSeekV3
GitHub - Rudra644/DeepSeek-Desktop-App: Experience DeepSeek's intelligent assistance instantly with this optimized desktop app. Enjoy faster performance, direct chat access, and a seamless, native exp...
Experience DeepSeek's intelligent assistance instantly with this optimized desktop app. Enjoy faster performance, direct chat access, and a seamless, native experience. - Rudra644/DeepSeek-Desk...
github.com
February 2, 2025 at 12:09 PM Everybody can reply
1 likes
1
srequena.bsky.social
Stéphane Requena
@srequena.bsky.social
👏🏻 @deepseek.bsky.social DeepseekV3 is out with SOTA performance beating Claude, Llama 405b etc on 10x less compute, $6m worth (2.8m hours) on 2048 H800s (cut down H100s) and trained using fp8 (quite rare even now on big LLM training)

github.com/deepseek-ai/...
GitHub - deepseek-ai/DeepSeek-V3
Contribute to deepseek-ai/DeepSeek-V3 development by creating an account on GitHub.
github.com
December 26, 2024 at 6:06 PM Everybody can reply
3 likes
3
qubite.bsky.social
Qubite
@qubite.bsky.social
Beware of the current top generative AI app"deepseek". It flagrantly violates ethics, funneling explicit content even from porn sites into outputs. We condemn this and will *never* tolerate such practices."
#DeepSeek #DeepSeekV3 #DeepSeek #DeepSeekR1 #deepseekaiagent
January 27, 2025 at 5:38 PM Everybody can reply
2 likes
2