Top Tweets for #eagi
A new https://t.co/KaEswSeHPA #EAGI (brains) 🔥🔥🔥 UPDATE, after the last deep review.
@OpenAI regains the leadership with GTP5.4 Pro Congrats, but also many other releases reshape the charts, and we welcome a new open 1Trillon model #Yuan 3.0 Ultra by @YuanAI_Lab
Remember that I test only native models not agregators, with custom real benchmarks:
My results as Mar 6th 2026:
🥇1st⬆️ OpenAI back to first with a great new model the GPT 5.4 Pro
1st Runner ↕️is @AnthropicAI barely keeping it with a still great 4.6 Opus model.
🥈2nd⬆️ @GoogleDeepMind also returns thanks to Gemini 3.1 Pro
2nd Runner is @Kimi_Moonshot #Kimi 2.5 great model for 🦞#openclaw @openclaw
3rd ↕️ @Zai_org keeps the position with the new model #GLM 5
3rd runer⬆️ @xai back to the game with new #Grok 4.2 mixture of 4 experts
4th⬇️ The previous revelation @StepFun_ai step-3.5-flash steps a few postions down but keeps being the best $ for 🦞
5th ↕️@Alibaba_Qwen keeps the postion with great new small models too.
6th ⬆️ @MiniMax_AI M2.5 keeps rising postions with a good model to have in the stack.
7th⬇️ @deepseek_ai the big promise steps down, we keep waiting for the next big one...
8th ↕️ @TheInclusionAI with #Ling & #Ring model are still good models but there is a lot of competition atm
8th runner a newcomer #YuanLab stay tunned this one looks promissing.
9th↕️ @Xiaomi #Mimo holds the postion
10th↕️ @Meituan_LongCat Flash Thinking is also a great model worth using.
Thanks to @huggingface for delivering the open models!
Learn more by joining https://t.co/KaEswSeHPA
(también en español en https://t.co/szKkt7tWpJ )
BTW this is this visible player list, stay tuned also for the work we are doing at @_Qubic_ #Neuraxon #Aigarth
#AI
#Artificiology
#ArtificiologyRanking
#Aritificiologia
#AGIRanking
#AGIRace
Time for a new https://t.co/KaEswSeHPA #EAGI (brains) 🔥🔥🔥 UPDATE, after the last deep review, 2026 is moving fast as expected.
First of all welcome to the new kid in the block @StepFun_ai with a triumphant #2 entry with step-3.5-flash a true killer and also the best price per token for your 🦞 #openclaw @openclaw thanks @steipete for releasing it! this model is free atm at @openrouter btw.
To de details: maybe @deepseek_ai will release V4 soon and also Gemini 3.1 is about to, but meanwhile, these are the current results:
Remember that I test only native models not agregators, with custom real benchmarks:
My results on Feb 12th 2026:
🥇1st⬆️ @AnthropicAI first time first with 4.6 Opus is killer model but expensive...
🥈2nd⬆️ The revelation @StepFun_ai step-3.5-flash also best $ for 🦞
3th⬆️ @Zai_org biggest raise with the great new model #GLM5
3rd (tie)⬇️ @Kimi_Moonshot #Kimi K2.5 Is a terrific new model
4th ⬇️ @GoogleDeepMind waiting for 3.1 and beyond...
4th (tie)⬇️ @OpenAI even if I agree the new codex 5.3 is delivering, the other models context size for use in chat app is just crap, will the former king lose the battle?
5th ⬇️ @xai #Grok is not there yet when a new good model?
5th (tie) ⬇️@Alibaba_Qwen are still great models still waiting for new ones
6th⬇️ @deepseek_ai also waiting for the next big one?
7th ⬆️ @MiniMax_AI M2.1 keeps rising 2.5 soon?
8th ⬆️ @TheInclusionAI with #Ling & #Ring model are still good models but there is a lot of competition atm
10th⬇️ @Xiaomi stil in the stage with #Mimo
10th↕️ @Meituan_LongCat Flash Thinking is also a great model worth using.
Thanks to @huggingface again for delivering the open models!
Learn more by joining https://t.co/KaEswSeHPA - Artificiology
(también en español en https://t.co/szKkt7tWpJ )
BTW this is this visible player list, stay tuned also for the work at @_Qubic_ #aigarth and #openscience #Neuraxon
#AI
#Artificiology
#ArtificiologyRanking
#Aritificiologia
#AGIRanking
#AGIRace
The full open version of the #EAGIBarometer by #Artificiology is available! Embodied #AGI or #EAGI is feeling closer than ever, I believe it's crucial to share & understand,scientifically, how close we really are, find out exploring the 144 barometer metrics. Find it https://t.co/KaEswSeHPA
https://t.co/KaEswSeHPA #EAGI (brains) 🔥🔥🔥 UPDATE, after the last deep review. (también en español en https://t.co/szKkt7tWpJ )
The list is getting crowded with 12 very close contenders in most of my latest tests. (some people ask what are the tests, basically what I call "edge-tests" things that have failed for me in the past (and most still fail btw) out of the regular known and learned benchmarks.
To de details: @OpenAI GPT5.2 Pro gets back to the top but shared for now with @GoogleDeepMind #Gemini3 Pro , but only thanks to GPT5.2 Pro the token limit in the others makes them unsuable for most heavy tasks unless you use the API.
Also new very interesting player enters the ranking @XiaomiMiMo with Mimo-V2-Flash, some people asked me about @MistralAI and it is true that it is improving but not reaching yet the list. Also remember that I test only native models not agregators, that mostly provide models show here.
My results on Dec 18th 2025:
🥇1st↕️ @GoogleDeepMind keeps the lead thanks to #Gemini3 pro but now shared
🥇1st ⬆️ @OpenAI jumps back to share first postion with #GPT 5.2 Pro
🥈2nd↕️@Kimi_Moonshot with #Kimi K2 Thinking keeps 2nd for now.
3rd⬆️@Alibaba_Qwen keep pushing the limits and rises to 3rd.
4th⬆️@deepseek_ai big jump with the new 3.2 models also best price atm.
5th↕️ @xai #Grok 4.1 keeps the position maybe something new by EOY?
6th↕️@AnthropicAI models are just good but besides coding not excelling yet.
7th ⬆️@minimax_ai M2 keeps rising
8th ⬆️@XiaomiMiMo enters the stage with #Mimo
9th⬇️ @TheInclusionAI with #Ling & #Ring models are still good but there is a lot of competition atm
10th↕️@Zai_org istill doing great with #GLM 4.6 and the new vision model
10th↕️@Meituan_LongCat Flash Thinking is also a great model worth using.
Thanks to @huggingface again for delivering the open models!
Learn more by joining https://t.co/KaEswSeHPA - Artificiology
#AI
#Artificiology
#ArtificiologyRanking
#Aritificiologia
#AGIRanking
#AGIRace
https://t.co/KaEswSeHPA #EAGI (brains) 🔥🔥🔥 UPDATE, after the last deep review. (también en español en https://t.co/szKkt7tWpJ )
Again worth noticing that being in the list is already a win since the 11 are very close in most of my tests, and indeed I use them all daily, I can say that from top to bottom there is only a 5% difference +- in my complex tasks benchmarks that I use to test them, and all the them lack of real memory beyond a few thousand tokens, and the output token limit is a burder for #LLMs atm to scale up.
@GoogleDeepMind with #Gemini3 conquers the top congrats @demishassabis and team, I would expect bigger jumps in performance, you know the current public benchmarks are not a good measure.
Also new very interesting player enters the ranking @minimax_ai , and for the first time @OpenAI steps down.
My results as Nov 20th 2025:
🥇1st⬆️ is @GoogleDeepMind Gemini 3.0 Pro is the current leader, for how long?
🥈2nd⬆️@Kimi_Moonshot with #Kimi K2 Thinking great jump and this is the best model for price and speed.
🥉3rd⬇️ @OpenAI sorry but GPT 5.1 Pro is not there yet, and the poor number of input tokens in ChatGPT app is limiting the potential.
4th⬇️@Alibaba_Qwen keep pushing the limits and probably will rise soon again.
5th↕️ is @xai #Grok new 4.1 is promissing but still instruction following is an issue to be solved.
6th⬇️ @AnthropicAI models are delivering but not excelling yet, and the limits in the app are ridiculous.
7th⬇️ is @TheInclusionAI with #Ling & #Ring model are promissing but thre is a lot of competition atm
8tk ⬆️@minimax_ai M2 has a great start a my bet is that next model will be closer to the top
9th⬇️ @deepseek_ai we are waiting for the next revolution?
10th⬇️ @Zai_org i is doing great with #GLM 4.6 but the fight is tight atm
10th↕️@Meituan_LongCat Flash Thinking is also a great model worth using.
Thanks to @huggingface for delivering the open models!
Learn more by joining https://t.co/KaEswSeHPA - Artificiology
#AI
#ArtificiologyRanking
#Aritificiologia
#AGIRanking
#AGIRace
https://t.co/KaEswSeHPA #EAGI (brains) 🔥🔥🔥 UPDATE, after the last deep review. (también en español en https://t.co/szKkt7tWpJ )
New format with 10 main players (removed the same spot ties) and added 3 cheapest "Intelligence" awards too 1 $ is the cheapest, and including @huggingface logo in the open weights/models (all the chinese btw)
Worth noticing that being in the list is already a win since the 10 are very close in most of my tests, and indeed I use them all daily.
A new very interesting player enters the ranking (also chinese) Ant International @TheInclusionAI Ling 1T and Ring 1T very worth using even if there seems to be some controversy with the metrics claims, models are at https://t.co/p6dYnBUoY8 and is still very new so I will do more tests to validate or challenge it.
My results on Oct 15th 2025:
🥇↕️ 1st @OpenAI thanks but barely to GPT 5 Pro (#Sora2 not part of the tests)
🥈⬆️ 2nd @Alibaba_Qwen keep pushing the limits with several great releases, Max is great and also love Tongyi results
🥉↕️ 3rd @AnthropicAI keeps the 3rd thanks to 4.5 model delivering
4th⬆️ is @TheInclusionAI with Ling1T (Ring 1T too) a incredible start
5th⬇️ is @xai Grok 4 Fast is great price/intelligence too but the new 10 slots moved to 5th but also wining the best 💸award in tie thanks to 4 Fast API prices.
6th⬇️ is @GoogleDeepMind delivering with 2.5 pro but wating for 3.0 to challenge the charts and maybe why not #genie3 world model release?
7th↕️ @deepseek_ai v.3.2 is a great model also wining the best 💸award in tie
8th↕️ @Zai_org is doing great with #GLM 4.6 but the fight is tight atm
9th↕️@Kimi_Moonshot with #Kimi K2 Instruct-0905 is still a favorite model of mine but there are many players
10th ↕️@Meituan_LongCat Flash Thinking is also a great model worth using.
Learn more by joining https://t.co/KaEswSeHPA - Artificiology
#AI #ArtificiologyRanking #Aritificiologia #AGIRanking #AGIRace
https://t.co/KaEswSeHPA #EAGI (brains) 🔥🔥🔥 UPDATE, after last deep review.
GPT 5 PRO still rules with Grok 4 Heavy, Qwen rises to second position thanks to Tongyi DeepResearch, and new commer LongCat Flash Thinking , Chinese open models keep pushing the field atm.
My results:
🥇↕️ 1st in tie @OpenAI thanks to GPT 5 Pro in tie
🥇↕️ 1st in tie @xai #Grok4 Heavy
🥈↕️ 2nd in tie @GoogleDeepMind 2.5 ultra is still delivering.
🥈⬆️ @Alibaba_Qwen Jumps to 2nd in tie, from 3th possition thanks to @Ali_TongyiLab
🥉↕️ @AnthropicAI still 3rd but a new #claude is needed soon to keep it.
🥉⬇️ @Kimi_Moonshot AI goes to third even with #Kim2 0905 is excelling, and one of my models of choice.
↕️ 4th @deepseek_ai V3.1 is good but when R2?
↕️ @Zai_org keeps 5th position thanks to #GLM 4.5
⬆️ @Meituan_LongCat enters the chart in 6th position thanks to #Longcat Flash Thinking
Learn more joining https://t.co/KaEswSeHPA - #Artificiology #AI
Humans & #Humanoids getting ready for the 1st Humanoid Olympiad in Olympia with @AcuminoAI @EndeavorGr and more. join https://t.co/dEVIxSJ6sG
Reading the https://t.co/PEOehakars #Trilogy to catch up #TheEndOfKnowledge #EAGI & #Artificracy
https://t.co/KaEswSeHPA #EAGI (brains) 🔥🔥🔥 UPDATE, after new deep review.
GPT 5 overpromissed in first tests results, besides a bit in coding, still no real long context in Chat, and SOTA specs improves are marginal, I will call it GPT 4.7 for now, Qwen new very long context is quite promissing and in general the Chinese open models are pushing the field atm.
Either the current #llm architectures are peeking or the main players are playing cat and mouse is very very small steps.
My results:
🥇↕️ 1st in tie @OpenAI thanks to GPT 5 Pro keeps the lead in tie
🥇↕️ 1st in tie @xai thanks #Grok4 Heavy
🥈↕️ 2nd in tie @GoogleDeepMind 2.5 ultra is delivering.
🥈⬆️ 2nd in tie @Kimi_Moonshot jumps again with #Kim2 excelling.
🥉↕️ @AnthropicAI jumps to 3rd #claude 4.1 and code helping.
🥉⬆️ @Alibaba_Qwen Jumps to 3rd from 5th possition, thanks to new models
⬇️ 4th from 2nd, @deepseek_ai when R2?
⬆️ @Zai_org enters the chart in 5th position thanks to #GLM 4.5
@Meta leaves the chart for now, let's see what they can do, besides calling Super Intelligence "Personal"....
Learn more joining https://t.co/KaEswSeHPA - #Artificiology #AI #AGI
After this month tests of the #EAGI (brains) 🔥🔥🔥 main players got a revolution in players with the first time @OpenAI 1st place is challenged by @xai and a new comer @Kimi_Moonshot with the first open model with 1T Parameters with great results jumping to 3rd position.
My results:
🥇↕️ 1st in tie @OpenAI
🥇⬆️ 1st in tie @xai thanks to #Grok4 what a comeback congrats @elonmusk
🥈↕️ 2nd in tie @deepseek_ai so far the best for the buck, but maybe not for longer.
🥈⬆️ 2nd in tie @GoogleDeepMind jumps again to main stage.
🥉⬆️ 3rd @Kimi_Moonshot gets to the list with the new #Kim2
↕️ @AnthropicAI keeps 4th with #claude 4 still with mixed results.
↕️ @Alibaba_Qwen keeps 5th possition.
↕️ @Meta keeps 6th position, for now the new hires will make the difference?
Learn more joining https://t.co/KaEswSeHPA - #Artificiology #AI
This month deep tests of the #EAGI (brains) 🔥🔥🔥 main players got a good shift in players with a big comeback of DeepSeek R1-0528 that is returing to second postion.
New Claude 4 versions are great at coding but underdelivers and rickety ratios made it unusable for many tasks, unless you pay the price, still waiting for Grok 3.5 maybe now that @elonmusk is back... June will be promissing!
My results:
🥇↕️ 1st is still @OpenAI but if price will be included in the mix DeepSeek will be first.
🥈⬆️ 2nd retaken by @deepseek_ai with the new R1-0528 just try it, so far the best for the buck.
🥉⬇️ @GoogleDeepMind even with the great Gemini 2.5 is challenged and beated by the new R1 in several aspects, besides multimodality and long context where Google still rules.
↕️ @AnthropicAI keeps 4th with the new claude 4 with mixed results.
⬇️ @xai #Grok3 goes down to 4th in tie
⬇️ @Alibaba_Qwen 3 goes down to 5th possition due to others going up and inconsistent results.
⬇️ @Meta goes down to 6th position, it is still in the list due to the long context / price.
Learn more joining https://t.co/KaEswSeHPA - #Artificiology
Coming home! @iclr_conf with 10K+ top #AI scientists in 😍 Singapur.
Shared ideas about #decentralizedAI, #bioinspiredAI, #Neuroscience #AGI, #EAGI with researchers and friend @VivancosDavid .
@alberfero Thank you for being the compass that guides @_Qubic_ through the tides.
First test of the just released @LumaLabsAI video 2 audio looks like #eagi is already developing language or anyone can translate? #DreamMachine
Looks like the #EAGI race is 🔥🔥🔥 main players shifted today, after a deep test of #grok3 my results:
🥇↕️ 1st is still @OpenAI
🥈@elonmusk ⬆️ to 2nd thanks to #Grok3
🥉@deepseek_ai ⬇️ to 3rd
4th tie between @AnthropicAI & @Google
Learn more at https://t.co/Ra8crOlrZT #Artificiology
🤖🎶#saturday #humanoid #robot #dance ? nice video @UnitreeRobotics
If you don't believe yet #EAGI Embodied Artificial General Intelligence is coming fast, you should learn more at https://t.co/PEOehakars #Artificiology #AI
@VivancosDavid @_Qubic_ Congratulations ! #EAGI is a main progress for humanity. The road to True #AI, decentralized for all humans, usable by everyone... Diamond project, dyor 💎 $Qubic #AGI
🚀Don't miss the #EAGI train....
Busy End of Year... leaving in a few hours to #Slovenia to keynote about @_Qubic_ #decentralized #TrueAI and present the premiere of my new 📘Embodied AGI (E-AGI) and The #Artificiology barometer🎯
And next week participating in one of my favorite #AI #ML 🎓conferences @NeurIPSConf this year with a fascinating set of 🥇keynote speakers like:
@AlisonGopnik @HochreiterSepp @drfeifei Lidong Zhou @ArnaudDoucet1 @DanicaKragic and @RosalindPicard a dozen #turorials 10's of #workshops and many more
👋Thanks @LesterMackey @amirgloberson @DaniCMBelg @jmtomczak @ulrichpaquet Angela Fan Cheng Zhang and many many others for organizing #Neurips 2024
And to the #Qubic team joining at #Ljubljana @albefero @foleycious @talentnodes @rok_straus @NejcZnidarcic and others that couldn't be there like @c___f___b @EricCLFung @josesanchezhb @JenniferChengLo and many more!
@CVLRwanda is transforming Rwanda’s housing sector through EAGI @EagiRwanda with innovative solutions that reduce costs and drive sustainable growth! Building homes, building the future.
#QualityHousing #SustainableGrowth
#EAGI
#RwOX
Last Seen Hashtags on Sotwe
nolimit() filter:videos
Seen from Germany
pinoybagets
Seen from Netherlands
xlii() or nolimit() or momson() or somno()
Seen from Mexico
JillianJanson
Seen from Turkey
bwwm
Seen from Latvia
breeding #sissy
Seen from Belarus
toiletslave
Seen from New Zealand
pasif
Seen from Turkey
เต้นเด้งควย
Seen from Thailand
deuleaks
Most Popular Users
Elon Musk 
@elonmusk
241.2M followers
Barack Obama
@barackobama
119.1M followers
Cristiano Ronaldo
@cristiano
112.8M followers
Donald J. Trump
@realdonaldtrump
111.8M followers
Narendra Modi
@narendramodi
107.1M followers
Rihanna
@rihanna
98.3M followers
NASA
@nasa
92.3M followers
Justin Bieber
@justinbieber
91.4M followers
KATY PERRY
@katyperry
89M followers
Taylor Swift
@taylorswift13
82.9M followers
Lady Gaga
@ladygaga
74.4M followers
Virat Kohli
@imvkohli
71.8M followers
Kim Kardashian
@kimkardashian
70.4M followers
YouTube
@youtube
68.8M followers
Neymar Jr
@neymarjr
64.8M followers
Bill Gates
@billgates
64.6M followers
The Ellen Show
@theellenshow
62.4M followers
Selena Gomez
@selenagomez
62M followers
CNN
@cnn
61.8M followers
X
@x
60.8M followers