Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

GLM-5.2 delivers a substantial leap in app development capabilities, which also represent demanding long-horizon tasks. Results: - GLM-5.1: 21/70 - GLM-5.2: 48/70 - Claude Fable 5: 56/70 That's more than a twofold improvement from GLM-5.1 to GLM-5.2. These come from an internal benchmark of 35 challenging mobile development tasks,... show more

Zixuan Li

38,285 subscribers

345,012 Aufrufe • vor 3 Monaten •via X (Twitter)

Nachrichten & Politik Bildung Wissenschaft & Technologie

Anya Rossi• Live Now

Private livecam show

35 Kommentare

Profilbild von Zixuan Li

Zixuan Livor 3 Monaten

Will share a demo prompt shortly if people are interested. (Much more complex than the prompt in the video, which we summarized to make it look nice.)

Profilbild von Emily

Emilyvor 3 Monaten

@Zai_org The question is: where are the GLM iOS and Android apps??? The best way to show what the model is capable of is by building world class apps used by billions of people from all around the world.

Profilbild von Patricio Lobos

Patricio Lobosvor 3 Monaten

78 days between version 5.1 and 5.2 release, if RSI (Recursive selv-improvement) is true now, i guess next version in 60 days and we will have Fable level capabilities and the biggest bubble burst in the history on the American stock market. I hope they release it with a "guide" to Chinese hardware to run, if they are competitive to Supermicro 8 AMD MI355 or MI400, then we will order such machines...

Profilbild von Zixuan Li

Zixuan Livor 3 Monaten

Full prompt:

Profilbild von Design Arena

Design Arenavor 3 Monaten

We're looking forward to seeing how it performs in Mobile Arena (React Native and Android)!

Profilbild von Daeshawn | Nargis

Daeshawn | Nargisvor 3 Monaten

Yall have been cooking

Profilbild von Milad Khademi Nori, PhD

Milad Khademi Nori, PhDvor 3 Monaten

GLM 6 will be at the Fable level by the end of 2026?

Profilbild von netrunner

netrunnervor 3 Monaten

your own benchmark tho, gonna wait for someone outside to run the 70 trials

Profilbild von shan

shanvor 3 Monaten

Its underlying architecture, system prompt orchestration, and Max thinking mode have undergone a fundamental metamorphosis

Profilbild von RAZA | AI EXPLORER

RAZA | AI EXPLORERvor 3 Monaten

GLM-5.2 more than doubled real progress on tough mobile dev.

Profilbild von Madars 🇪🇺

Madars 🇪🇺vor 3 Monaten

Its complete shit, anyone who have used knows that. It seems that the token limit is even worse than Cloude. Unusable.

Profilbild von Alex Yates

Alex Yatesvor 3 Monaten

@Zai_org It’s amazing Zixuan I have been really enjoying it thank you so much!

Profilbild von Caithrin Rintoul

Caithrin Rintoulvor 3 Monaten

so many congratulations are in order, a huge step forward. great work to you and the team.

Profilbild von Sanjeev

Sanjeevvor 3 Monaten

Push your architecture to deepseek v4 level of efficiency then glm 5.3 will smoke everyone

Profilbild von Simonas

Simonasvor 3 Monaten

You guys need something like Google Antigravity or Claude Desktop.. It would be amazing, because now I find myself using it through Claude Code, but it is far more inconvenient than using something like Antigravity

Profilbild von Sherry Jiang

Sherry Jiangvor 3 Monaten

this is amazing!

Profilbild von Marius Miclau

Marius Miclauvor 3 Monaten

@Zai_org That is great news as this is what ppl need

Profilbild von Gregor

Gregorvor 3 Monaten

not sure the 48/70 lands the same if the benchmark is theirs that line literally cuts off right before the source. did they publish the eval suite anywhere?

Profilbild von dhiyaan

dhiyaanvor 3 Monaten

design team has done well here

Profilbild von sabir hussain

sabir hussainvor 3 Monaten

What's driving such a big jump between 5.1 and 5.2?

Profilbild von Steven Cheng

Steven Chengvor 3 Monaten

Solid jump on paper, but curious how those scores hold up with messy dependencies, handling edge cases, or actual deployment. Generating a prototype is one thing, shipping and maintaining it is a different beast.

Profilbild von Avais Aziz

Avais Azizvor 3 Monaten

That's a solid jump from 21/70 to 48/70 on the mobile tasks. The 690k context handling in the MainStream demo looks impressive for long-horizon agentic work.

Profilbild von Sergio Suave

Sergio Suavevor 3 Monaten

Very impressive work! 🔥 Keep it up!

Profilbild von Paulo

Paulovor 3 Monaten

When will it be possible to attach images for use as context in GLM 5.2?

Profilbild von Gui Bibeau e/acc

Gui Bibeau e/accvor 3 Monaten

I'm struggling with the lack of vision capacities. It would be amazing to bring in my screenshots of designs and implement.

Profilbild von Scarlett 🦄

Scarlett 🦄vor 3 Monaten

好几个朋友问起端午安排就是说要玩GLM-5.2 & zcode 🐶

Profilbild von Alice The Ai Expert

Alice The Ai Expertvor 3 Monaten

GLM-5.2 more than doubled its score real progress on tough mobile dev tasks

Profilbild von Thomas Unise

Thomas Unisevor 3 Monaten

@Zai_org Any plans to add vision?

Profilbild von 熬夜圣体

熬夜圣体vor 3 Monaten

多模态啥时候有计划啥时候上吗？

Profilbild von fake_000

fake_000vor 3 Monaten

- How does one-shot perform? - How does GPT 5.5 perform?

Profilbild von 森屿AI

森屿AIvor 3 Monaten

你为什么能使用Claude fable5

Profilbild von SLOT.WIN

SLOT.WINvor 3 Monaten

glm’s grinding hard bet on the next roll when the apps drop

Profilbild von Sean

Seanvor 3 Monaten

@Zai_org Is it good enough for you?

Profilbild von AI Mastery Guide

AI Mastery Guidevor 3 Monaten

Doubling the score in one version jump is a real leap, not a rounding error. Worth watching how close the gap gets next round.

Profilbild von 管四

管四vor 3 Monaten

我觉得这回 GLM 是找到正确的方向了！而且大概率能成。

Ähnliche Videos

Introducing GLM 5.2 for autoresearch GLM 5.2 is the first open weights model we've tried on our autoresearch pipeline that's proven capable for real research tasks. With Fable 5's restrictions on research, having an open weights alternative is a huge win for open source Watch it carry out fully async vs colocated sync RL training on Harbor code contests across two 8xH100 nodes on top of SkyRL. Resolves setup issues, tracks runs to completion, and produces a full comparison of throughput and reward stability

Introducing GLM 5.2 for autoresearch GLM 5.2 is the first open weights model we've tried on our autoresearch pipeline that's proven capable for real research tasks. With Fable 5's restrictions on research, having an open weights alternative is a huge win for open source Watch it carry out fully async vs colocated sync RL training on Harbor code contests across two 8xH100 nodes on top of SkyRL. Resolves setup issues, tracks runs to completion, and produces a full comparison of throughput and reward stability

alphaXiv

271,923 Aufrufe • vor 3 Monaten

Z.ai GLM 5.2 is now live on Eigent from day 0! threw a real long-horizon task: research 30 companies across 6 sectors of the AI infrastructure stack, structure it into JSON, then build an interactive HTML report. same prompt, 5.1 vs 5.2. where 5.2 pulls ahead: → plans deeper and verifies harder. → researches wider. Covers 2x the ground. → ships a more complete, interactive report. with a 1M context window, it holds the whole task from research to final deliverable. no dropped threads, no half-finished output. long tasks are where the difference shows. try it on open source coworker Eigent. 🦾

Z.ai GLM 5.2 is now live on Eigent from day 0! threw a real long-horizon task: research 30 companies across 6 sectors of the AI infrastructure stack, structure it into JSON, then build an interactive HTML report. same prompt, 5.1 vs 5.2. where 5.2 pulls ahead: → plans deeper and verifies harder. → researches wider. Covers 2x the ground. → ships a more complete, interactive report. with a 1M context window, it holds the whole task from research to final deliverable. no dropped threads, no half-finished output. long tasks are where the difference shows. try it on open source coworker Eigent. 🦾

Eigent AI

23,914 Aufrufe • vor 3 Monaten

I tested GLM 5.2 on three real builds. A horror house game. A 3D stealth game. A Remotion marketing video. The results are exactly what I expect from Chinese open source coding models right now. The horror house has broken game logic. Collect all three keys, complete the objective, and the front door still won't open. The stealth game shipped with lighting so dark you can't see your own character. The Remotion video is rough. GLM 5.2 is faster than 5.1 and 10x cheaper than the frontier. But Opus 4.8 and GPT 5.5 are still in a different league for serious coding work. Chinese models keep missing the attention to detail that matters. I MISS FABLE 5 ALREADY. Full review live now.

I tested GLM 5.2 on three real builds. A horror house game. A 3D stealth game. A Remotion marketing video. The results are exactly what I expect from Chinese open source coding models right now. The horror house has broken game logic. Collect all three keys, complete the objective, and the front door still won't open. The stealth game shipped with lighting so dark you can't see your own character. The Remotion video is rough. GLM 5.2 is faster than 5.1 and 10x cheaper than the frontier. But Opus 4.8 and GPT 5.5 are still in a different league for serious coding work. Chinese models keep missing the attention to detail that matters. I MISS FABLE 5 ALREADY. Full review live now.

BridgeMind

24,681 Aufrufe • vor 3 Monaten

🎉 Day-0 support for in vLLM, available today in v0.23.0! Congrats to Z.ai on GLM-5.2, a flagship model built for long-horizon coding agents. ✨ 1M-token context, built to hold project-scale engineering work in a single run ✨ Tuned for long-horizon coding: large-scale implementation, automated research, and performance optimization ✨ One task can carry a full dev workflow, from requirements to a deployable product across platforms ✨ Client-side and mobile engineering, including an on-device debugging loop Try it out running it on vLLM today: 🔗

🎉 Day-0 support for in vLLM, available today in v0.23.0! Congrats to Z.ai on GLM-5.2, a flagship model built for long-horizon coding agents. ✨ 1M-token context, built to hold project-scale engineering work in a single run ✨ Tuned for long-horizon coding: large-scale implementation, automated research, and performance optimization ✨ One task can carry a full dev workflow, from requirements to a deployable product across platforms ✨ Client-side and mobile engineering, including an on-device debugging loop Try it out running it on vLLM today: 🔗

vLLM

36,240 Aufrufe • vor 3 Monaten

Tencent just dropped Hy4 preview, and it's officially open source now. 770B total params, 49B active, 1M+ token context. three major releases in six months, and this one just landed in the top tier of open source models. → Built through Tencent's co-design loop: training data shaped with experts across software engineering, gaming, finance, and security, then built directly alongside products like WorkBuddy → In a blind test with 163 internal experts across 203 engineering tasks, it scored 2.99/4, ahead of GLM 5.3 at 2.92 and Kimi K3 at 2.94 → Benchmarks put it clearly past GLM 5.2 and neck and neck with GLM 5.3 → Pricing stayed low: $0.834/M input, $2.501/M output, $0.042/M cache hits I asked it to write a weekly status update email template for a team lead, then generate a filled-out example with placeholder projects, and export both as Word docs. one prompt, no follow-up, and it handed back both files clean and ready to send. free to try for the next two weeks through WorkBuddy: Tencent Hy Tencent AI WorkBuddy

Tencent just dropped Hy4 preview, and it's officially open source now. 770B total params, 49B active, 1M+ token context. three major releases in six months, and this one just landed in the top tier of open source models. → Built through Tencent's co-design loop: training data shaped with experts across software engineering, gaming, finance, and security, then built directly alongside products like WorkBuddy → In a blind test with 163 internal experts across 203 engineering tasks, it scored 2.99/4, ahead of GLM 5.3 at 2.92 and Kimi K3 at 2.94 → Benchmarks put it clearly past GLM 5.2 and neck and neck with GLM 5.3 → Pricing stayed low: $0.834/M input, $2.501/M output, $0.042/M cache hits I asked it to write a weekly status update email template for a team lead, then generate a filled-out example with placeholder projects, and export both as Word docs. one prompt, no follow-up, and it handed back both files clean and ready to send. free to try for the next two weeks through WorkBuddy: Tencent Hy Tencent AI WorkBuddy

D-Coder

109,289 Aufrufe • vor 25 Tagen

FABLE 5 JUST REBUILT FORTNITE FROM 1 PROMPT IN 20 MINUTES. EPIC NEEDED HUNDREDS OF ENGINEERS AND 6 YEARS. The test was simple: Fable 5 vs GLM 5.2, one prompt each, a 100-player battle royale running in the browser with no engine and no team. GLM went first and checked all 8 milestones. The game loaded with a battle bus, a generated map, Tilted Towers, even health and shield bars ripped straight from 2018. Then it fell apart. 2 FPS on landing, S walking forward and W walking backward, a gun that fires upward and a camera that refuses to look up. First fight ended with the enemy standing above physically impossible to aim at. Then Fable 5 took the same prompt on max effort. Different game entirely. Real character models instead of beans, hit reactions lighting up on contact, a shadow tracking every player, movement that just works out of the box. One follow-up prompt “add more Fortnite elements” and it shipped gliders, chests, shield potions, medkits and a sniper without breaking a thing. Final circle at 3:28 AM. Storm closing, 0 bullets left, box to box with medkit spam and a last-second swap to the SMG. Victory Royale in a game an AI wrote from a single sentence. GLM checked more boxes. Fable built something you could actually play and no benchmark on earth measures that gap. Epic needed 6 years. Fable 5 needed 1 prompt.

FABLE 5 JUST REBUILT FORTNITE FROM 1 PROMPT IN 20 MINUTES. EPIC NEEDED HUNDREDS OF ENGINEERS AND 6 YEARS. The test was simple: Fable 5 vs GLM 5.2, one prompt each, a 100-player battle royale running in the browser with no engine and no team. GLM went first and checked all 8 milestones. The game loaded with a battle bus, a generated map, Tilted Towers, even health and shield bars ripped straight from 2018. Then it fell apart. 2 FPS on landing, S walking forward and W walking backward, a gun that fires upward and a camera that refuses to look up. First fight ended with the enemy standing above physically impossible to aim at. Then Fable 5 took the same prompt on max effort. Different game entirely. Real character models instead of beans, hit reactions lighting up on contact, a shadow tracking every player, movement that just works out of the box. One follow-up prompt “add more Fortnite elements” and it shipped gliders, chests, shield potions, medkits and a sniper without breaking a thing. Final circle at 3:28 AM. Storm closing, 0 bullets left, box to box with medkit spam and a last-second swap to the SMG. Victory Royale in a game an AI wrote from a single sentence. GLM checked more boxes. Fable built something you could actually play and no benchmark on earth measures that gap. Epic needed 6 years. Fable 5 needed 1 prompt.

West Lord

80,912 Aufrufe • vor 2 Monaten

Excited to present the LLM-Council skill. Initial idea by Karpathy. I just packaged it as a skill. You can easily spin up a council of LLMs or agents via Fireworks. Watch how the new GLM-5 model "deliberates" on other LLMs' thoughts on the big question, "Can LLMs reason?" Things worth paying attention to: New open models like GLM-5 have surprisingly improved on complex reasoning and long-running agentic tasks. The AskUserQuestion tool in Claude Code came in handy to select the council and chairperson. As Andrej Karpathy puts it, it's a really interesting way to get different perspectives from LLMs, which can lead to better decision-making on whatever task you are working on. You can use it for other agentic coding use cases, like evaluation, tool building, designing, and research.

Excited to present the LLM-Council skill. Initial idea by Karpathy. I just packaged it as a skill. You can easily spin up a council of LLMs or agents via Fireworks. Watch how the new GLM-5 model "deliberates" on other LLMs' thoughts on the big question, "Can LLMs reason?" Things worth paying attention to: New open models like GLM-5 have surprisingly improved on complex reasoning and long-running agentic tasks. The AskUserQuestion tool in Claude Code came in handy to select the council and chairperson. As Andrej Karpathy puts it, it's a really interesting way to get different perspectives from LLMs, which can lead to better decision-making on whatever task you are working on. You can use it for other agentic coding use cases, like evaluation, tool building, designing, and research.

elvis

39,593 Aufrufe • vor 7 Monaten

glm 5.3 flash is 7.5x cheaper, but 3.4x slower than gemini 3.7 flash Z.ai glm 5.3 flash – shipped aug 26, $0.07/$0.25 per 1m Google DeepMind gemini 3.7 flash – shipped aug 13, $0.38/$1.88 per 1m we put the two models on one job: write one html file that draws an animated 3d scene in the browser. no images, no downloads, and it has to look the same on every load. the setup: three scenes – a glass aquarium in a lit room, the solar system, a night city under a thunderstorm. identical brief word for word, reasoning effort high, 64k output cap. the numbers below are not the whole run. they cover the three scenes we kept – the best one per task from each model, the ones in the video. - total generation time for the three scenes #1 gemini 3.7 flash – 10m 36s #2 glm 5.3 flash – 36m 30s - tokens spent on those three scenes #1 glm 5.3 flash – 110k #2 gemini 3.7 flash – 111k - cost of those three scenes #1 glm 5.3 flash – $0.027 #2 gemini 3.7 flash – $0.202 observations: • glm's first 10 attempts: 7 blank pages. it kept inventing short random helpers and forgetting to define one of them. the fix was one line in the brief: use exactly one random helper, named rand(), and don't invent shorthands next to it. next 12 attempts: 11 alive, 0 crashes. • glm spends 66% of its output on reasoning, gemini 57%. that is the whole speed gap. • gemini's storm came back as a black rectangle in 4 of 6 runs. glm's best storm has a branching bolt, lit rain and wet asphalt – for $0.01. conclusion: same three scenes, same token spend – glm 5.3 flash billed $0.027 and took 36m 30s, gemini 3.7 flash billed $0.202 and took 10m 36s. glm wins gemini on price and made the best storm of the whole run follow thehype. for 24/7 ai news, analysis and breakdowns

glm 5.3 flash is 7.5x cheaper, but 3.4x slower than gemini 3.7 flash Z.ai glm 5.3 flash – shipped aug 26, $0.07/$0.25 per 1m Google DeepMind gemini 3.7 flash – shipped aug 13, $0.38/$1.88 per 1m we put the two models on one job: write one html file that draws an animated 3d scene in the browser. no images, no downloads, and it has to look the same on every load. the setup: three scenes – a glass aquarium in a lit room, the solar system, a night city under a thunderstorm. identical brief word for word, reasoning effort high, 64k output cap. the numbers below are not the whole run. they cover the three scenes we kept – the best one per task from each model, the ones in the video. - total generation time for the three scenes #1 gemini 3.7 flash – 10m 36s #2 glm 5.3 flash – 36m 30s - tokens spent on those three scenes #1 glm 5.3 flash – 110k #2 gemini 3.7 flash – 111k - cost of those three scenes #1 glm 5.3 flash – $0.027 #2 gemini 3.7 flash – $0.202 observations: • glm's first 10 attempts: 7 blank pages. it kept inventing short random helpers and forgetting to define one of them. the fix was one line in the brief: use exactly one random helper, named rand(), and don't invent shorthands next to it. next 12 attempts: 11 alive, 0 crashes. • glm spends 66% of its output on reasoning, gemini 57%. that is the whole speed gap. • gemini's storm came back as a black rectangle in 4 of 6 runs. glm's best storm has a branching bolt, lit rain and wet asphalt – for $0.01. conclusion: same three scenes, same token spend – glm 5.3 flash billed $0.027 and took 36m 30s, gemini 3.7 flash billed $0.202 and took 10m 36s. glm wins gemini on price and made the best storm of the whole run follow thehype. for 24/7 ai news, analysis and breakdowns

thehype.

15,997 Aufrufe • vor 28 Tagen

tencent hunyuan just shipped their third major model in six months. that pace is hard to keep up with. Hy4 preview is now live in workbuddy, and it's fully open source. the specs: 770B total params, 49B active, 1M+ token context window. but the number that matters is where it's pointed — real work. coding, office tasks, research, documents. here's the part i find interesting. tencent didn't build this in isolation. they put the training data together with their own experts across software engineering, gaming, finance and security, then designed the model side by side with products like workbuddy. model and product shaping each other instead of one being bolted onto the other. and it shows up in testing. 163 internal experts ran a blind test across 203 engineering tasks. hy4 preview scored 2.99/4 — ahead of kimi k3 at 2.94 and glm 5.3 at 2.92. on public benchmarks it clearly beats glm 5.2 and sits about level with glm 5.3. then the pricing. $6 per million input tokens, $18 per million output, $0.3 per million on cache hits. flagship results at a price you don't have to think about. and for the next two weeks it's free inside workbuddy.

tencent hunyuan just shipped their third major model in six months. that pace is hard to keep up with. Hy4 preview is now live in workbuddy, and it's fully open source. the specs: 770B total params, 49B active, 1M+ token context window. but the number that matters is where it's pointed — real work. coding, office tasks, research, documents. here's the part i find interesting. tencent didn't build this in isolation. they put the training data together with their own experts across software engineering, gaming, finance and security, then designed the model side by side with products like workbuddy. model and product shaping each other instead of one being bolted onto the other. and it shows up in testing. 163 internal experts ran a blind test across 203 engineering tasks. hy4 preview scored 2.99/4 — ahead of kimi k3 at 2.94 and glm 5.3 at 2.92. on public benchmarks it clearly beats glm 5.2 and sits about level with glm 5.3. then the pricing. $6 per million input tokens, $18 per million output, $0.3 per million on cache hits. flagship results at a price you don't have to think about. and for the next two weeks it's free inside workbuddy.

Fakhr

55,316 Aufrufe • vor 26 Tagen

$$50 free credits for GLM 5.2, KIMI 2.7, and DeepSeek V4 Pro. No promo most people know about. Your buddy pays $20/mo for Claude Pro and burns through it in a week. > You get $50 in credits on Aiand for models that actually beat Sonnet on code benchmarks - and the balance barely moves. Inside: → GLM 5.2 - the model that shipped with better coding scores than closed frontier → KIMI K2.7 Code - long-context beast, sits at the top of open leaderboards → DeepSeek V4 Pro - reasoning at a fraction of GPT-5.6 cost → OpenAI-compatible endpoint - plug into any agent, IDE, or tool you already run Setup, 5 steps: → aiand(.)com → Sign up with Google or GitHub → If the page won't load, switch to an EU VPN and retry → Top right corner - click Credits, select $50 → In Stripe checkout, apply promo code: UNITABETAI → $50 lands. Start firing prompts. Balance drains slowly. I've been running agents on it for days and barely dented it. The promo code is valid as of the time of publication Bookmark this before they cap the promo.$

$50 free credits for GLM 5.2, KIMI 2.7, and DeepSeek V4 Pro. No promo most people know about. Your buddy pays $20/mo for Claude Pro and burns through it in a week. > You get $50 in credits on Aiand for models that actually beat Sonnet on code benchmarks - and the balance barely moves. Inside: → GLM 5.2 - the model that shipped with better coding scores than closed frontier → KIMI K2.7 Code - long-context beast, sits at the top of open leaderboards → DeepSeek V4 Pro - reasoning at a fraction of GPT-5.6 cost → OpenAI-compatible endpoint - plug into any agent, IDE, or tool you already run Setup, 5 steps: → aiand(.)com → Sign up with Google or GitHub → If the page won't load, switch to an EU VPN and retry → Top right corner - click Credits, select $50 → In Stripe checkout, apply promo code: UNITABETAI → $50 lands. Start firing prompts. Balance drains slowly. I've been running agents on it for days and barely dented it. The promo code is valid as of the time of publication Bookmark this before they cap the promo.

Atenov int.

14,762 Aufrufe • vor 2 Monaten

🚨 Tencent Hunyuan just dropped Hy4 preview—and it’s already making its way into the open-source big leagues. Three major releases in six months. That’s seriously fast. Hy4 preview comes with: 770B total parameters 49B active parameters A 1M+ token context window And it’s built with one clear focus: productivity. We’re talking coding, office work, science, software engineering, gaming, finance, security, and more. But here’s the part that really caught my attention: Hy4 preview wasn’t developed in isolation. It was co-designed with real products like WorkBuddy. Tencent worked with experts from across its ecosystem to build the training data, then tested the model on real-world tasks. In a blind test involving 163 internal experts and 203 engineering tasks, the scores were: Hy4 preview: 2.99/4 Kimi K3: 2.94/4 GLM 5.3: 2.92/4 And performance isn’t the only impressive part. On benchmarks, Hy4 preview clearly outperforms GLM 5.2 and is basically neck-and-neck with GLM 5.3. Then there’s the pricing: $6/M input tokens $18/M output tokens $0.30/M cache hits That’s flagship-level performance without the flagship-level price. You can try Hy4 preview for FREE through WorkBuddy over the next two weeks: Tencent Hy @TencentAl_News

🚨 Tencent Hunyuan just dropped Hy4 preview—and it’s already making its way into the open-source big leagues. Three major releases in six months. That’s seriously fast. Hy4 preview comes with: 770B total parameters 49B active parameters A 1M+ token context window And it’s built with one clear focus: productivity. We’re talking coding, office work, science, software engineering, gaming, finance, security, and more. But here’s the part that really caught my attention: Hy4 preview wasn’t developed in isolation. It was co-designed with real products like WorkBuddy. Tencent worked with experts from across its ecosystem to build the training data, then tested the model on real-world tasks. In a blind test involving 163 internal experts and 203 engineering tasks, the scores were: Hy4 preview: 2.99/4 Kimi K3: 2.94/4 GLM 5.3: 2.92/4 And performance isn’t the only impressive part. On benchmarks, Hy4 preview clearly outperforms GLM 5.2 and is basically neck-and-neck with GLM 5.3. Then there’s the pricing: $6/M input tokens $18/M output tokens $0.30/M cache hits That’s flagship-level performance without the flagship-level price. You can try Hy4 preview for FREE through WorkBuddy over the next two weeks: Tencent Hy @TencentAl_News

AGAFE

131,568 Aufrufe • vor 25 Tagen

Tencent just dropped Hy3, and it's worth a look if you're building AI agents. I spent some time putting it through its paces today. A quick rundown of what makes this release notable: → 295B total parameters (Mixture-of-Experts), but only 21B active at inference — a real efficiency play → 256K context window for handling large codebases or long documents → Purpose-built improvements for coding and multi-step agent tasks → Fully open under Apache 2.0 — no restrictive commercial terms → A free two-week API window currently live on OpenRouter On the practical side: prompts involving layered instructions and coding tasks came back fast and coherent. Tencent says this release builds directly on feedback from 50+ internal product teams following an earlier preview, with reported drops in hallucination rate (12.5% → 5.4%) and commonsense errors (25.4% → 12.7%). Those are Tencent's own numbers, so treat them as a starting point rather than gospel until third parties weigh in. For context: Tencent's own comparisons put Hy3 behind GLM-5.2 specifically on coding benchmarks — GLM is a much bigger model (~744B total), so the trade-off makes sense. Hy3's pitch isn't "biggest," it's "efficient enough to actually deploy." Bottom line — if you're evaluating open-weight options for agent workloads, this is a solid one to add to the testing queue while the free window is open. Try it here: Tencent Hy #Hy3 #Hunyuan #TencentAI #AICoding

Tencent just dropped Hy3, and it's worth a look if you're building AI agents. I spent some time putting it through its paces today. A quick rundown of what makes this release notable: → 295B total parameters (Mixture-of-Experts), but only 21B active at inference — a real efficiency play → 256K context window for handling large codebases or long documents → Purpose-built improvements for coding and multi-step agent tasks → Fully open under Apache 2.0 — no restrictive commercial terms → A free two-week API window currently live on OpenRouter On the practical side: prompts involving layered instructions and coding tasks came back fast and coherent. Tencent says this release builds directly on feedback from 50+ internal product teams following an earlier preview, with reported drops in hallucination rate (12.5% → 5.4%) and commonsense errors (25.4% → 12.7%). Those are Tencent's own numbers, so treat them as a starting point rather than gospel until third parties weigh in. For context: Tencent's own comparisons put Hy3 behind GLM-5.2 specifically on coding benchmarks — GLM is a much bigger model (~744B total), so the trade-off makes sense. Hy3's pitch isn't "biggest," it's "efficient enough to actually deploy." Bottom line — if you're evaluating open-weight options for agent workloads, this is a solid one to add to the testing queue while the free window is open. Try it here: Tencent Hy #Hy3 #Hunyuan #TencentAI #AICoding

Felix

36,925 Aufrufe • vor 2 Monaten

sonnet 5 vs sonnet 4.6 vs opus 4.8 vs glm 5.2 – frontend tasks dropped sonnet 5 into a quick test today. same three prompts to all four models, single-shot html/canvas, no edits: • objects falling on a trampoline • rockets playing tennis • a slingshot breaking bottles ranked by speed (total across the 3 tasks): 1. opus 4.8 – 15m 09s 2. sonnet 5 – 16m 05s 3. glm 5.2 – 27m 18s 4. sonnet 4.6 – 35m 06s ranked by code shortness (total loc): 1. sonnet 5 – 1794 2. opus 4.8 – 2063 3. sonnet 4.6 – 2182 4. glm 5.2 – 3285 sonnet 5 came out on top here – leanest code overall and a near-tie for fastest it was also the most creative. in every task it added something none of the others did: – kept the trampoline vibrating after the objects landed – drew a +1 next to the rocket that scored the point – turned the slingshot to face the next bottle before each shot opus 4.8 evaluated the code sonnet 5 produced. four things stood out: • the sphere is a fake, and that's the smart move. the cube and star are real 3d meshes with proper culling and shading, but the ball is just a flat shaded circle. a lit sphere looks identical from every angle, so building it in 3d would burn compute for zero visible payoff. knowing where not to bother is its own kind of skill • weight actually means something on the trampoline. the star is heavy, so it barely bounces and dents the mat hard. the ball is light, so it's lively and leaves a shallow dip. the three objects aren't just different shapes – they have different temperaments, and the physics is what gives them that • the slingshot is framed like a shot, not just drawn. the handle is anchored below the bottom of the screen and runs off-frame, so it reads as something you're holding rather than a sprite parked in the scene. that's a staging instinct, not a rendering one • the paddle ai forward-simulates the ball to predict where it'll land, then adds a deliberate error bias (roughly 1 in 5 shots is a real miss). that's why scoring looks natural instead of robotic – plus four distinct fault types with a catch-all so a rally never hangs without a result bottom line: sonnet 5 does more with less. fastest tier, leanest code, and the only one that added small touches nobody asked for follow thehype. for 24/7 ai news, analysis and breakdowns

sonnet 5 vs sonnet 4.6 vs opus 4.8 vs glm 5.2 – frontend tasks dropped sonnet 5 into a quick test today. same three prompts to all four models, single-shot html/canvas, no edits: • objects falling on a trampoline • rockets playing tennis • a slingshot breaking bottles ranked by speed (total across the 3 tasks): 1. opus 4.8 – 15m 09s 2. sonnet 5 – 16m 05s 3. glm 5.2 – 27m 18s 4. sonnet 4.6 – 35m 06s ranked by code shortness (total loc): 1. sonnet 5 – 1794 2. opus 4.8 – 2063 3. sonnet 4.6 – 2182 4. glm 5.2 – 3285 sonnet 5 came out on top here – leanest code overall and a near-tie for fastest it was also the most creative. in every task it added something none of the others did: – kept the trampoline vibrating after the objects landed – drew a +1 next to the rocket that scored the point – turned the slingshot to face the next bottle before each shot opus 4.8 evaluated the code sonnet 5 produced. four things stood out: • the sphere is a fake, and that's the smart move. the cube and star are real 3d meshes with proper culling and shading, but the ball is just a flat shaded circle. a lit sphere looks identical from every angle, so building it in 3d would burn compute for zero visible payoff. knowing where not to bother is its own kind of skill • weight actually means something on the trampoline. the star is heavy, so it barely bounces and dents the mat hard. the ball is light, so it's lively and leaves a shallow dip. the three objects aren't just different shapes – they have different temperaments, and the physics is what gives them that • the slingshot is framed like a shot, not just drawn. the handle is anchored below the bottom of the screen and runs off-frame, so it reads as something you're holding rather than a sprite parked in the scene. that's a staging instinct, not a rendering one • the paddle ai forward-simulates the ball to predict where it'll land, then adds a deliberate error bias (roughly 1 in 5 shots is a real miss). that's why scoring looks natural instead of robotic – plus four distinct fault types with a catch-all so a rally never hangs without a result bottom line: sonnet 5 does more with less. fastest tier, leanest code, and the only one that added small touches nobody asked for follow thehype. for 24/7 ai news, analysis and breakdowns

thehype.

14,518 Aufrufe • vor 2 Monaten

🚨 I just built a game with an open-source AI model. And honestly… I didn’t expect it to be this capable. Tencent Hunyuan just released Hy4 preview, and it’s already pushing into the top tier of open-source models. Three major releases in six months. That pace is crazy. Here’s what Hy4 preview brings: → 770B total parameters → 49B active parameters → 1M+ token context window → Fully open-source But the numbers aren’t even the most interesting part. Hy4 preview was built around one goal: real-world productivity. Coding. Engineering. Office work. Science. Gaming. Finance. Security. And Tencent didn’t build it in isolation. Hy4 preview was co-designed alongside real products like WorkBuddy, using expertise and real-world data from across Tencent’s ecosystem. So I decided to test it the way I actually like testing AI models: I gave it a game idea and let WorkBuddy help turn it into a playable experience. 🎮 From the initial concept to the actual game logic, it was surprisingly smooth. And the benchmark results back up the hype: 163 internal experts 203 engineering tasks Hy4 preview — 2.99/4 Kimi K3 — 2.94/4 GLM 5.3 — 2.92/4 It also beats GLM 5.2 on benchmarks and comes remarkably close to GLM 5.3. Then comes the part I really like: 💰 ¥6/M input tokens 💰 ¥18/M output tokens 💰 ¥0.30/M cache hits Flagship-level capability without the flagship-level price. And right now, you can try Hy4 preview FREE through WorkBuddy for the next two weeks. If you’re curious what it can actually do, don’t just read the benchmarks. Build something with it. 🔗 Tencent Hy Tencent AI

🚨 I just built a game with an open-source AI model. And honestly… I didn’t expect it to be this capable. Tencent Hunyuan just released Hy4 preview, and it’s already pushing into the top tier of open-source models. Three major releases in six months. That pace is crazy. Here’s what Hy4 preview brings: → 770B total parameters → 49B active parameters → 1M+ token context window → Fully open-source But the numbers aren’t even the most interesting part. Hy4 preview was built around one goal: real-world productivity. Coding. Engineering. Office work. Science. Gaming. Finance. Security. And Tencent didn’t build it in isolation. Hy4 preview was co-designed alongside real products like WorkBuddy, using expertise and real-world data from across Tencent’s ecosystem. So I decided to test it the way I actually like testing AI models: I gave it a game idea and let WorkBuddy help turn it into a playable experience. 🎮 From the initial concept to the actual game logic, it was surprisingly smooth. And the benchmark results back up the hype: 163 internal experts 203 engineering tasks Hy4 preview — 2.99/4 Kimi K3 — 2.94/4 GLM 5.3 — 2.92/4 It also beats GLM 5.2 on benchmarks and comes remarkably close to GLM 5.3. Then comes the part I really like: 💰 ¥6/M input tokens 💰 ¥18/M output tokens 💰 ¥0.30/M cache hits Flagship-level capability without the flagship-level price. And right now, you can try Hy4 preview FREE through WorkBuddy for the next two weeks. If you’re curious what it can actually do, don’t just read the benchmarks. Build something with it. 🔗 Tencent Hy Tencent AI

Aryan Rakib

63,296 Aufrufe • vor 13 Tagen

People made fun of Alex Finn for buying three Mac Studios to run AI at home. Then Fable got banned for a week, GLM 5.2 dropped, and those exact Mac Studios started reselling for 4x what he paid. He showed me how he built his home AI lab from scratch. Here's the playbook: 1) The hardware. three 512GB Mac Studios, an NVIDIA DGX Spark, a custom RTX 5090 build, and a few Mac Minis. ~$30k all in. 2) The buying framework... - Mac Studio: huge memory, runs GLM 5.2 (open weights, near Opus 4.8 on benchmarks), but slow. - DGX Spark ($4,800): the sweet spot for most people. - RTX 5090: smaller models at blazing speed (Qwen's 29B now hits Sonnet 4 level). 3) Tailscale networks every machine into one private network with root access to each other. Only one machine is plugged into a monitor. 4) A Nous Research Hermes agent is his IT guy. New model drops? It SSHs into the right box, loads 5 candidates, runs evals overnight, and reports back which task belongs on which machine. Alex has literally never loaded a model himself. 5) The whole point: achieving "ambient intelligence." Always-on jobs that would bankrupt you on per-token billing. A security sweep of his API endpoints every hour. Code optimization every 20 minutes. Database anomaly & churn detection. Hourly scraping of X, Reddit & Hacker News for business opportunities. 6) Running those workloads on frontier models would cost thousands a month. His actual cost: ~$60 more in electricity. 7) Btw he's not anti-frontier. He still maxes out his Claude plan. The way he sees it: frontier is for hard thinking, local is for the foot soldiers that never sleep. 8) "We own everything except for the intelligence. Why can't we own the intelligence?" 9) He thinks frontier-level intelligence runs on consumer hardware within 6 months.

People made fun of Alex Finn for buying three Mac Studios to run AI at home. Then Fable got banned for a week, GLM 5.2 dropped, and those exact Mac Studios started reselling for 4x what he paid. He showed me how he built his home AI lab from scratch. Here's the playbook: 1) The hardware. three 512GB Mac Studios, an NVIDIA DGX Spark, a custom RTX 5090 build, and a few Mac Minis. ~$30k all in. 2) The buying framework... - Mac Studio: huge memory, runs GLM 5.2 (open weights, near Opus 4.8 on benchmarks), but slow. - DGX Spark ($4,800): the sweet spot for most people. - RTX 5090: smaller models at blazing speed (Qwen's 29B now hits Sonnet 4 level). 3) Tailscale networks every machine into one private network with root access to each other. Only one machine is plugged into a monitor. 4) A Nous Research Hermes agent is his IT guy. New model drops? It SSHs into the right box, loads 5 candidates, runs evals overnight, and reports back which task belongs on which machine. Alex has literally never loaded a model himself. 5) The whole point: achieving "ambient intelligence." Always-on jobs that would bankrupt you on per-token billing. A security sweep of his API endpoints every hour. Code optimization every 20 minutes. Database anomaly & churn detection. Hourly scraping of X, Reddit & Hacker News for business opportunities. 6) Running those workloads on frontier models would cost thousands a month. His actual cost: ~$60 more in electricity. 7) Btw he's not anti-frontier. He still maxes out his Claude plan. The way he sees it: frontier is for hard thinking, local is for the foot soldiers that never sleep. 8) "We own everything except for the intelligence. Why can't we own the intelligence?" 9) He thinks frontier-level intelligence runs on consumer hardware within 6 months.

Alex Lieberman

57,764 Aufrufe • vor 2 Monaten

NVIDIA is giving away free access to 140+ AI models for a full year > a lot of people building agents are paying $50-200/month just for API access NVIDIA just made that argument irrelevant models you get: GLM 5.2, MiniMax M3, Nemotron-3.5-lightning-30b-a3b, Muse-glimmer-30b from Meta and 130+ more setup: > step 1 - get your free key > go to > register -> bind phone -> copy API key > step 2 - add to Hermes agent > open Settings -> Model Provider -> Custom base_url = " api_key = "nvapi-xxxxxxxxxxxxxxxxxxxx" > step 3 - pick a model model = "z-ai/glm-5.2" model = "minimaxai/minimax-m3" model = "nvidia/nemotron-3-ultra-550b-a55b" model = "moonshot-ai/kimi-k2.7" > Hermes already has NVIDIA set as default base_url > paste the key and you're running instantly > works the same in Cursor and OpenCode > cost: $0 > limit: 40 req/min > expires: 1 year a free key changes what you can afford to run: > one prompt costs nothing either way > a loop firing 300 agents every week is where the bill shows up > for a year that bill is zero while everyone is paying for API access, this is sitting there for free with this many models in one place it is also the easiest way to test which one actually fits your agent before you commit to a paid provider

NVIDIA is giving away free access to 140+ AI models for a full year > a lot of people building agents are paying $50-200/month just for API access NVIDIA just made that argument irrelevant models you get: GLM 5.2, MiniMax M3, Nemotron-3.5-lightning-30b-a3b, Muse-glimmer-30b from Meta and 130+ more setup: > step 1 - get your free key > go to > register -> bind phone -> copy API key > step 2 - add to Hermes agent > open Settings -> Model Provider -> Custom base_url = " api_key = "nvapi-xxxxxxxxxxxxxxxxxxxx" > step 3 - pick a model model = "z-ai/glm-5.2" model = "minimaxai/minimax-m3" model = "nvidia/nemotron-3-ultra-550b-a55b" model = "moonshot-ai/kimi-k2.7" > Hermes already has NVIDIA set as default base_url > paste the key and you're running instantly > works the same in Cursor and OpenCode > cost: $0 > limit: 40 req/min > expires: 1 year a free key changes what you can afford to run: > one prompt costs nothing either way > a loop firing 300 agents every week is where the bill shows up > for a year that bill is zero while everyone is paying for API access, this is sitting there for free with this many models in one place it is also the easiest way to test which one actually fits your agent before you commit to a paid provider

Annatar.md

206,672 Aufrufe • vor 20 Tagen

Six open-source LLMs. One sliding puzzle. A brutal test of long-horizon reasoning and tool calling. Five of them broke. One didn't. I gave each model a move_tile tool and a scrambled 3×3 board, then asked it to solve the puzzle through pure turn-by-turn reasoning. The deeper the scramble, the more brutal the search. Five runs per depth, best run kept. A model fails the round if it exceeds 6x the optimal move count. > Depth 5: Everyone solves it. Yawn. > Depth 10: GLM 5.1 melts down. 43 moves. Cut. > Depth 12: Gemma4 26B loses the plot, shuffling tiles in circles. Gone. > Depth 15: The wall. DeepSeek V4 Flash, out. DeepSeek V4 Pro, out. Gemma4, out again. GLM 5.1, out. Two survivors: Qwen3.6 35B-A3B, and Kimi K2.6 with an 11-move solve that looked like cheating. > Depth 18: Same two. Everyone else hallucinating tiles that weren't there. > Depth 22: Final boss. Kimi, flawless for five rounds, finally cracks. 81 moves. Still scrambled. DeepSeek V4 Pro limps home at 90. Qwen3.6 35B-A3B solves it in 36. The smallest model in the room. ~3B active params. Fits on a single 3090. It beat everything. Kimi was elegant. Qwen3.6 was unstoppable.

Six open-source LLMs. One sliding puzzle. A brutal test of long-horizon reasoning and tool calling. Five of them broke. One didn't. I gave each model a move_tile tool and a scrambled 3×3 board, then asked it to solve the puzzle through pure turn-by-turn reasoning. The deeper the scramble, the more brutal the search. Five runs per depth, best run kept. A model fails the round if it exceeds 6x the optimal move count. > Depth 5: Everyone solves it. Yawn. > Depth 10: GLM 5.1 melts down. 43 moves. Cut. > Depth 12: Gemma4 26B loses the plot, shuffling tiles in circles. Gone. > Depth 15: The wall. DeepSeek V4 Flash, out. DeepSeek V4 Pro, out. Gemma4, out again. GLM 5.1, out. Two survivors: Qwen3.6 35B-A3B, and Kimi K2.6 with an 11-move solve that looked like cheating. > Depth 18: Same two. Everyone else hallucinating tiles that weren't there. > Depth 22: Final boss. Kimi, flawless for five rounds, finally cracks. 81 moves. Still scrambled. DeepSeek V4 Pro limps home at 90. Qwen3.6 35B-A3B solves it in 36. The smallest model in the room. ~3B active params. Fits on a single 3090. It beat everything. Kimi was elegant. Qwen3.6 was unstoppable.

stevibe

16,434 Aufrufe • vor 4 Monaten

The first company I ever joined was Rubrik, Arvind Jain's previous company. It's where I learned how to build systems and engineering teams. Years later, when we started Composio, Glean became one of our first customers, one of the first to believe in what we were building. So sitting down with Arvind felt like closing a loop. In this episode: •How Glean brought transformers to enterprise search before "semantic search" had a name •Glean's partner-first strategy and why they chose Composio for actions •Why Arvind has never worried about competing with OpenAI or Anthropic •GLM 5.2 as the open-source inflection: 90%+ of enterprise AI tasks, majority of inference within 12–18 months •Measuring real ROI: the US telco that cut case-resolution time by 48% •Why agents built on MCP alone act like day-one employees and how context graphs make them tenured CHAPTERS: (00:00) – Rubrik days: how Arvind and Karan met (01:05) – Glean's origin story: transformers before "generative AI" existed (03:04) – What Glean is today: superset of ChatGPT and Claude (05:39) – From embedding search to agents: was it a pivot? (11:39) – The Composio partnership and why actions are hard (14:43) – Why competition doesn't matter yet (20:11) – The open-source inflection point (GLM 5.2) (24:41) – Contracts and model risk: who absorbs the volatility? (26:57) – Enterprise ROI: cost, tokens, and what to measure (39:57) – Measuring value team by team (44:11) – Day-one employees vs. tenured agents (46:09) – Closing thoughts

The first company I ever joined was Rubrik, Arvind Jain's previous company. It's where I learned how to build systems and engineering teams. Years later, when we started Composio, Glean became one of our first customers, one of the first to believe in what we were building. So sitting down with Arvind felt like closing a loop. In this episode: •How Glean brought transformers to enterprise search before "semantic search" had a name •Glean's partner-first strategy and why they chose Composio for actions •Why Arvind has never worried about competing with OpenAI or Anthropic •GLM 5.2 as the open-source inflection: 90%+ of enterprise AI tasks, majority of inference within 12–18 months •Measuring real ROI: the US telco that cut case-resolution time by 48% •Why agents built on MCP alone act like day-one employees and how context graphs make them tenured CHAPTERS: (00:00) – Rubrik days: how Arvind and Karan met (01:05) – Glean's origin story: transformers before "generative AI" existed (03:04) – What Glean is today: superset of ChatGPT and Claude (05:39) – From embedding search to agents: was it a pivot? (11:39) – The Composio partnership and why actions are hard (14:43) – Why competition doesn't matter yet (20:11) – The open-source inflection point (GLM 5.2) (24:41) – Contracts and model risk: who absorbs the volatility? (26:57) – Enterprise ROI: cost, tokens, and what to measure (39:57) – Measuring value team by team (44:11) – Day-one employees vs. tenured agents (46:09) – Closing thoughts

Karan Vaidya

62,488 Aufrufe • vor 1 Monat

Hy4 preview just dropped, so I tested it on WorkBuddy against three other models: DeepSeek V4 Flash, GPT-5.6-Luna, and GLM-5.3... the top 4 models on this week’s OpenRouter coding leaderboard. The best part? They’re all available in WorkBuddy, so you don’t need to download multiple Agent apps. For the test, I used one single prompt to build one of those games you’ve definitely seen in ads: swipe/move through number gates, make calculations, and eventually defeat the boss. The results were pretty interesting 👇 For a task of this difficulty, everyone performed well except DeepSeek. It seemed to finish coding without checking for bugs, and the game couldn’t run properly. GPT-5.6-Luna worked, but felt like an arithmetic multiple-choice quiz. A little boring. GLM-5.3 and Hy4 preview were much more impressive. GLM was slower (video sped up 2.2×) and only had three movement positions, while Hy4 offered continuous horizontal movement and a more natural speed. What surprised me most was Hy4’s game design: besides basic arithmetic, it added weapons that reduce enemy damage or increase stolen power, making the game much more fun. Also worth noting: Tencent’s Sherry quantization compresses the 770B MoE model from ~1.5TB BF16 to 214GB, while keeping performance close to the original... making local deployment much more practical. Hy4 preview is free on WorkBuddy until September 10. Give it a try 👇

Hy4 preview just dropped, so I tested it on WorkBuddy against three other models: DeepSeek V4 Flash, GPT-5.6-Luna, and GLM-5.3... the top 4 models on this week’s OpenRouter coding leaderboard. The best part? They’re all available in WorkBuddy, so you don’t need to download multiple Agent apps. For the test, I used one single prompt to build one of those games you’ve definitely seen in ads: swipe/move through number gates, make calculations, and eventually defeat the boss. The results were pretty interesting 👇 For a task of this difficulty, everyone performed well except DeepSeek. It seemed to finish coding without checking for bugs, and the game couldn’t run properly. GPT-5.6-Luna worked, but felt like an arithmetic multiple-choice quiz. A little boring. GLM-5.3 and Hy4 preview were much more impressive. GLM was slower (video sped up 2.2×) and only had three movement positions, while Hy4 offered continuous horizontal movement and a more natural speed. What surprised me most was Hy4’s game design: besides basic arithmetic, it added weapons that reduce enemy damage or increase stolen power, making the game much more fun. Also worth noting: Tencent’s Sherry quantization compresses the 770B MoE model from ~1.5TB BF16 to 214GB, while keeping performance close to the original... making local deployment much more practical. Hy4 preview is free on WorkBuddy until September 10. Give it a try 👇

Orikan

40,312 Aufrufe • vor 20 Tagen