Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

Round 2: Ming-Image-0.1-Design vs Qwen Image 2.1! This time in ComfyUI - same exact prompt, identical seed, with an auto stitched side-by-side. Full prompt in comments. - Both got all 4 captions right. - Ming: a detailed illustration for every step, like a real cafe poster. - Qwen was... show more

Mia

38,636 subscribers

10,304 görüntüleme • 10 gün önce •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

15 Yorum

Mia profil fotoğrafı

Mia10 gün önce

Prompt used in this video: Clean flat infographic titled "POUR-OVER IN 4 STEPS". Four numbered panels with simple icons and captions: "1 RINSE THE FILTER", "2 ADD 20G COFFEE", "3 BLOOM 30 SECONDS", "4 POUR 320G WATER". Beige background, brown accents, modern layout. HF links:

HealthRanger profil fotoğrafı

HealthRanger10 gün önce

Very cool, I'm loading weights and gonna try this myself on a 6000 pro, with side by side compared to qwen-image-2.1. Will share results.

James Malsawm profil fotoğrafı

James Malsawm10 gün önce

This look very impressive

JJ Eaton profil fotoğrafı

JJ Eaton10 gün önce

Wow. Ming is amazing. How does it perform with editing existing images?

💫 Dor 🎗️ profil fotoğrafı

💫 Dor 🎗️10 gün önce

Who is you? great account @MiaAI_lab

Soveryn Intelligence profil fotoğrafı

Soveryn Intelligence10 gün önce

Round 2 makes the split clear — Ming takes the cafe-poster detail, Qwen wins wall-clock 35s vs 90s on the 5090 at matched seed. Did you lock steps/CFG too, or was seed the only fixed knob?

BullBear.News profil fotoğrafı

BullBear.News10 gün önce

ComfyUI node execution order on the image loader step still drops metadata if you pass custom dicts through native loaders.

Don profil fotoğrafı

Don10 gün önce

ming looking pretty good. Looking forward for the barrage of tests you will put it through.

_alphashark_ profil fotoğrafı

_alphashark_10 gün önce

in ComfyUI, pin the sampler and scheduler, not just the seed. I’d also rerun at each model’s native resolution since latent rescaling can skew detail retention.

LocalAiLab profil fotoğrafı

LocalAiLab10 gün önce

Hmm... Intressting. I will try it. Thx for the workflow

Sanjeev profil fotoğrafı

Sanjeev9 gün önce

You post this kind of tweets also??? Awesome dud. I can get to experience this in less time.well Appreciated

Ekithanun profil fotoğrafı

Ekithanun10 gün önce

The side-by-side setup is such a smart test—great to see both models handle the prompt so well.

Mike Hutu profil fotoğrafı

Mike Hutu10 gün önce

Wow ming seems like a great model ! Need to load the weights!

Mia profil fotoğrafı

Mia10 gün önce

Try both. It's better than Qwen in some scenarios and worse in others.

Mike Hutu profil fotoğrafı

Mike Hutu10 gün önce

I’ve been using ideogram4 and it has been a very good model for high quality images (mostly ugc type i’ve tried) and the json prompting + bbox mapping is really great feature to get the objects exactly where i want them to be. I’ll download both of these. Been really busy lately..

Benzer Videolar

🚀 New update for the Grafilab AI ecosystem! We’ve just updated new frontier models in our Playground and Unified API, including the highly anticipated models by DeepSeek and Qwen families! 🧠⚡ But that’s not all… you can now run Side-by-Side Model Comparisons directly in the Grafilab Playground! 🥊 Compare outputs from Gemini, DeepSeek, Qwen, and more in real-time, on the exact same prompt, with zero code. Find the perfect model for your workload instantly. Try it out now:

🚀 New update for the Grafilab AI ecosystem! We’ve just updated new frontier models in our Playground and Unified API, including the highly anticipated models by DeepSeek and Qwen families! 🧠⚡ But that’s not all… you can now run Side-by-Side Model Comparisons directly in the Grafilab Playground! 🥊 Compare outputs from Gemini, DeepSeek, Qwen, and more in real-time, on the exact same prompt, with zero code. Find the perfect model for your workload instantly. Try it out now:

Grafilab - Unleash AI

805,542 görüntüleme • 3 ay önce

Today, we open-source Pruna-Qwen-Image-2.1, a set of a few-step LoRA adapters that make Qwen-Image-2.1 by Qwen up to 6.3× faster for image generation and editing. - 𝗛𝗼𝘄 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁 𝗶𝘀 𝗶𝘁? Generate or edit images in just 5 or 8 steps instead of 40. Choose 5 steps for maximum speed or 8 steps for the best balance. - 𝗛𝗼𝘄 𝗴𝗼𝗼𝗱 𝗶𝘀 𝗶𝘁? The 8-step adapter is our recommended default. It supports text-to-image and single- or multi-image editing at 1K resolution, with up to three reference images. - 𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗶𝘁 𝘄𝗼𝗿𝗸?The LoRA adapters load directly on top of Qwen-Image-2.1, while the pipeline remains unchanged. Each adapter uses its own optimized sigma schedule and runs without CFG. 𝘛𝘩𝘪𝘴 𝘪𝘴 𝘫𝘶𝘴𝘵 𝘰𝘶𝘳 𝘧𝘪𝘳𝘴𝘵 𝘳𝘦𝘭𝘦𝘢𝘴𝘦, 𝘸𝘪𝘵𝘩 𝘦𝘷𝘦𝘯 𝘣𝘦𝘵𝘵𝘦𝘳 𝘷𝘦𝘳𝘴𝘪𝘰𝘯𝘴 𝘰𝘯 𝘵𝘩𝘦 𝘸𝘢𝘺. 🚀 Want to try the fast open-source integration? Check it here on Diffusers: Want to try the fastest image generation and editing endpoint? Check P-Image-Ideogram and P-Image-Edit here on API:

Today, we open-source Pruna-Qwen-Image-2.1, a set of a few-step LoRA adapters that make Qwen-Image-2.1 by Qwen up to 6.3× faster for image generation and editing. - 𝗛𝗼𝘄 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁 𝗶𝘀 𝗶𝘁? Generate or edit images in just 5 or 8 steps instead of 40. Choose 5 steps for maximum speed or 8 steps for the best balance. - 𝗛𝗼𝘄 𝗴𝗼𝗼𝗱 𝗶𝘀 𝗶𝘁? The 8-step adapter is our recommended default. It supports text-to-image and single- or multi-image editing at 1K resolution, with up to three reference images. - 𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗶𝘁 𝘄𝗼𝗿𝗸?The LoRA adapters load directly on top of Qwen-Image-2.1, while the pipeline remains unchanged. Each adapter uses its own optimized sigma schedule and runs without CFG. 𝘛𝘩𝘪𝘴 𝘪𝘴 𝘫𝘶𝘴𝘵 𝘰𝘶𝘳 𝘧𝘪𝘳𝘴𝘵 𝘳𝘦𝘭𝘦𝘢𝘴𝘦, 𝘸𝘪𝘵𝘩 𝘦𝘷𝘦𝘯 𝘣𝘦𝘵𝘵𝘦𝘳 𝘷𝘦𝘳𝘴𝘪𝘰𝘯𝘴 𝘰𝘯 𝘵𝘩𝘦 𝘸𝘢𝘺. 🚀 Want to try the fast open-source integration? Check it here on Diffusers: Want to try the fastest image generation and editing endpoint? Check P-Image-Ideogram and P-Image-Edit here on API:

Pruna AI

90,322 görüntüleme • 20 gün önce

Alright guys, wow. Qwen 3.8 27B is basically Qwen 4. What we have here is one of the best one-shot shark survival games I've made with any model. I ran this same prompt through fable, and the biggest thing all models struggle with in this test is the top-down view of the sharks. Fable even had all of the fins reversed and details jumbled on the first turn. Before now, it was really a 2 to 8 turn test to get something final regardless of model. Qwen 3.8 27B just nailed this in one output, no harness at all, just purely the model thoroughly thinking before outputting the final result. On every asset used, different shark variants, everything, it all looks incredibly tight. The boat movement, shark speed and engagement, everything is so much better than I expected, and I was expecting a lot. Imagine what 8 turns would do in a harness? This confirms to me that while Qwen 3.8 27B thinks a lot, for the first time in a local model on the base weights, IT MIGHT ACTUALLY BE WORTH IT This is a Frontier Lab model result, and it ran on my 5090 at 60tps in Q5_K_M. I am just in complete astoundment at what Qwen has accomplished here, and I am incredibly grateful that this model is open source. The speed will also increase drastically with optimisations. This is just an incredible day. What a blessing. More to come!

Alright guys, wow. Qwen 3.8 27B is basically Qwen 4. What we have here is one of the best one-shot shark survival games I've made with any model. I ran this same prompt through fable, and the biggest thing all models struggle with in this test is the top-down view of the sharks. Fable even had all of the fins reversed and details jumbled on the first turn. Before now, it was really a 2 to 8 turn test to get something final regardless of model. Qwen 3.8 27B just nailed this in one output, no harness at all, just purely the model thoroughly thinking before outputting the final result. On every asset used, different shark variants, everything, it all looks incredibly tight. The boat movement, shark speed and engagement, everything is so much better than I expected, and I was expecting a lot. Imagine what 8 turns would do in a harness? This confirms to me that while Qwen 3.8 27B thinks a lot, for the first time in a local model on the base weights, IT MIGHT ACTUALLY BE WORTH IT This is a Frontier Lab model result, and it ran on my 5090 at 60tps in Q5_K_M. I am just in complete astoundment at what Qwen has accomplished here, and I am incredibly grateful that this model is open source. The speed will also increase drastically with optimisations. This is just an incredible day. What a blessing. More to come!

Kyle Hessling

26,198 görüntüleme • 2 ay önce

ONE PROMPT IN CHATGPT IMAGE 2 JUST REPLACED AN ENTIRE $5,000-$50,000 BRAND DESIGN TEAM Full brand identity. Colors, typography, logo style, aesthetic. All generated in one single prompt. Here's the exact workflow: Step 1: Paste one prompt asking for a complete brand identity. GPT Image 2 generates everything – color palette, fonts, logo direction, overall visual style. Step 2: Use that output as context for every asset you generate after. Campaign visuals, shop graphics, product mockups, social posts – all stay consistent and on-brand automatically. The result: A full visual identity that looks like a real design team built it. Clean text rendering. Consistent style across every single asset. Design agencies charge $5,000-$50,000 for brand identity work. GPT Image 2 does it in one prompt. The full prompt that makes this work is in the video below. Bookmark this post, you'll want to come back to it.

SCOTTY BEAM

12,023 görüntüleme • 3 ay önce

GLM-5.3 Flash vs Qwen3.8 Flash-Next. Both local on my 4x 3090, both left alone with 1 prompt per task. 11 scenes, 18 hours, 2.4M tokens out. All of it is in the video. I wanted a clear winner. I didnt get one, but I know which one stays on my rig. GLM-5.3 Flash: Maya-M quant, 116 GB, runs in Project Maya on all 4 cards plus RAM. 35 to 48 tok/s while it works. Qwen3.8 Flash-Next: IQ3_S quant, 78 GB, runs in Strata on 2 cards. 85 to 118 tok/s. Both have 262K context and vision, so the agent opens its own page, takes a screenshot, looks at it and fixes what it sees. 1 prompt per scene. No follow ups and no hints from me. 120 minutes max per scene, 150 for the kart game. I ran them in 2 harnesses, OMP first and then DeepSeek Harness. Minutes until the model said it was done: Pelican in OMP: GLM 39, Qwen 66 Pelican in DeepSeek Harness: GLM 93, Qwen 86 Kart game in OMP: GLM 130, Qwen stopped by the 150 limit Millipede in OMP: both stopped by the 120 limit Millipede in DeepSeek Harness: Qwen stopped at 120 again Battle diorama in DeepSeek Harness: GLM 81, Qwen 113 Thats 11 of 16 combos. I ran out of time for the other 5, sorry. They work in a very different way and I didnt expect that. GLM thinks for a long time and then writes a lot at once. Its first pelican took only 36 turns. For the diorama it gave 1 answer of 47K tokens that ran for almost 18 minutes, with the full scene in a single file. Qwen goes in small steps and checks everything. The same pelican was 240 turns and 45 browser checks. In total GLM wrote 513K tokens in 5 runs and Qwen wrote 1.9M in 6 runs, so Qwen puts out 2.6x more per minute of real time, on half the cards. The harness also changed more than I thought it would. OMP keeps only 5 images in a request when the provider is local. So every new screenshot removes the oldest one, the prompt changes in the middle, and the server has to read the context again from the start. GLM lost 37 of its 130 kart minutes just on that. In DeepSeek Harness, on the diorama, it lost 3.5 minutes out of 81. Qwen reads a prompt almost 4x faster on my rig (84K tokens in 23 seconds, GLM needs 100 seconds for 96K), so in OMP it lost only 9 to 12 minutes per run. I cant give you a winner. GLM is the only one that finished the kart game inside the limit, and its scenes look cleaner to me. But its millipede is bad. Qwen puts more detail into everything, and its diorama and millipede from DeepSeek Harness are my favorite scenes in the video. It also got stopped by the time limit in 3 of its 6 runs, because it just keeps polishing. For my own use I'd take Qwen. It is a lot faster and it needs only 2 cards, so the other 2 stay free and I can run something else next to it. This is 1 run per combo, so dont read it as statistics. Both models ran with my refusal projection on, because thats how I use them every day. And in the kart clips the game's own AI is driving the player kart, I only recorded the screen. Music: YuE2

GLM-5.3 Flash vs Qwen3.8 Flash-Next. Both local on my 4x 3090, both left alone with 1 prompt per task. 11 scenes, 18 hours, 2.4M tokens out. All of it is in the video. I wanted a clear winner. I didnt get one, but I know which one stays on my rig. GLM-5.3 Flash: Maya-M quant, 116 GB, runs in Project Maya on all 4 cards plus RAM. 35 to 48 tok/s while it works. Qwen3.8 Flash-Next: IQ3_S quant, 78 GB, runs in Strata on 2 cards. 85 to 118 tok/s. Both have 262K context and vision, so the agent opens its own page, takes a screenshot, looks at it and fixes what it sees. 1 prompt per scene. No follow ups and no hints from me. 120 minutes max per scene, 150 for the kart game. I ran them in 2 harnesses, OMP first and then DeepSeek Harness. Minutes until the model said it was done: Pelican in OMP: GLM 39, Qwen 66 Pelican in DeepSeek Harness: GLM 93, Qwen 86 Kart game in OMP: GLM 130, Qwen stopped by the 150 limit Millipede in OMP: both stopped by the 120 limit Millipede in DeepSeek Harness: Qwen stopped at 120 again Battle diorama in DeepSeek Harness: GLM 81, Qwen 113 Thats 11 of 16 combos. I ran out of time for the other 5, sorry. They work in a very different way and I didnt expect that. GLM thinks for a long time and then writes a lot at once. Its first pelican took only 36 turns. For the diorama it gave 1 answer of 47K tokens that ran for almost 18 minutes, with the full scene in a single file. Qwen goes in small steps and checks everything. The same pelican was 240 turns and 45 browser checks. In total GLM wrote 513K tokens in 5 runs and Qwen wrote 1.9M in 6 runs, so Qwen puts out 2.6x more per minute of real time, on half the cards. The harness also changed more than I thought it would. OMP keeps only 5 images in a request when the provider is local. So every new screenshot removes the oldest one, the prompt changes in the middle, and the server has to read the context again from the start. GLM lost 37 of its 130 kart minutes just on that. In DeepSeek Harness, on the diorama, it lost 3.5 minutes out of 81. Qwen reads a prompt almost 4x faster on my rig (84K tokens in 23 seconds, GLM needs 100 seconds for 96K), so in OMP it lost only 9 to 12 minutes per run. I cant give you a winner. GLM is the only one that finished the kart game inside the limit, and its scenes look cleaner to me. But its millipede is bad. Qwen puts more detail into everything, and its diorama and millipede from DeepSeek Harness are my favorite scenes in the video. It also got stopped by the time limit in 3 of its 6 runs, because it just keeps polishing. For my own use I'd take Qwen. It is a lot faster and it needs only 2 cards, so the other 2 stay free and I can run something else next to it. This is 1 run per combo, so dont read it as statistics. Both models ran with my refusal projection on, because thats how I use them every day. And in the kart clips the game's own AI is driving the player kart, I only recorded the screen. Music: YuE2

Alexey Fateev

12,116 görüntüleme • 3 gün önce

I know it's hard to believe, but this IQ_3 of Qwen 3.8 Flash Next is genuinely impressive. Sure, it's not BF16, but it works very well in Claude Code. To eliminate the likelihood of benchmaxxing, I present a novel voxel benchmark: Voxel Sermon on the Mount. This was generated with the ISTA Labs Qwen 3.8 Flash next IQ_3S quant at 150 tps in Claude Code, running on a single rtx 5090. Niko Veit's Astra engine for the hybrid speed! No loops, no garbled tool calls; it was just done in a few minutes. No image head used; it built this without seeing it.

I know it's hard to believe, but this IQ_3 of Qwen 3.8 Flash Next is genuinely impressive. Sure, it's not BF16, but it works very well in Claude Code. To eliminate the likelihood of benchmaxxing, I present a novel voxel benchmark: Voxel Sermon on the Mount. This was generated with the ISTA Labs Qwen 3.8 Flash next IQ_3S quant at 150 tps in Claude Code, running on a single rtx 5090. Niko Veit's Astra engine for the hybrid speed! No loops, no garbled tool calls; it was just done in a few minutes. No image head used; it built this without seeing it.

Kyle Hessling

11,680 görüntüleme • 14 gün önce

Round 2: Dense edition Qwen3.5 27B vs Gemma4 31B Same 4 HTML canvas challenges, this time no MoE. Setup: > Quant: Unsloth Q6_K_XL (both) > Each running on an RTX 5090 > Single HTML file, no libraries, real-time animation My read per prompt: 🌳 Tree — Gemma, branches are beautifully structured ✨ Night sky — close, but Qwen's stars blink better 🐟 Fish boids — Qwen, detailed fish and look at that motion, so realistic 🏠 House build — neither really nailed it Agree? Does dense actually beat MoE here?

Round 2: Dense edition Qwen3.5 27B vs Gemma4 31B Same 4 HTML canvas challenges, this time no MoE. Setup: > Quant: Unsloth Q6_K_XL (both) > Each running on an RTX 5090 > Single HTML file, no libraries, real-time animation My read per prompt: 🌳 Tree — Gemma, branches are beautifully structured ✨ Night sky — close, but Qwen's stars blink better 🐟 Fish boids — Qwen, detailed fish and look at that motion, so realistic 🏠 House build — neither really nailed it Agree? Does dense actually beat MoE here?

stevibe

33,304 görüntüleme • 5 ay önce

Qwen 3.8 27B beat Claude Opus 4.6 for $0 open-weights Qwen 3.8 27B on one rented GPU vs anthropic's Claude Opus 4.6. same prompt, no edits the test: one self-contained three.js file rendering sunken Atlantis — temple, columns, boid fish, kelp, bubbles, god rays, caustics — plus a 20-second camera flythrough cost: Qwen 3.8 27B — free (open weights, self-hosted) Claude Opus 4.6 — $0.21 what we saw: • qwen actually built a temple — full colonnade, domed roof, glowing portal centered. opus made a long row of columns fading into fog • qwen's portal is the star of the shot. opus leaves it small and off to the side • qwen packed more into frame — fish, kelp, ruins layered together. opus spreads thin and empty qwen's weights are on hugging face — you can run this one yourself!

Qwen 3.8 27B beat Claude Opus 4.6 for $0 open-weights Qwen 3.8 27B on one rented GPU vs anthropic's Claude Opus 4.6. same prompt, no edits the test: one self-contained three.js file rendering sunken Atlantis — temple, columns, boid fish, kelp, bubbles, god rays, caustics — plus a 20-second camera flythrough cost: Qwen 3.8 27B — free (open weights, self-hosted) Claude Opus 4.6 — $0.21 what we saw: • qwen actually built a temple — full colonnade, domed roof, glowing portal centered. opus made a long row of columns fading into fog • qwen's portal is the star of the shot. opus leaves it small and off to the side • qwen packed more into frame — fish, kelp, ruins layered together. opus spreads thin and empty qwen's weights are on hugging face — you can run this one yourself!

AI/ML API

102,960 görüntüleme • 2 ay önce

Here's a comparison i never thought i would make Qwen-3.8-27b vs Gemini-3.7-flash Yes, you read it right! I don't know what to say, this is soo damn impressive. the difference is hugeee🤯 Take a look at this result 😍 (watch it in highest quality) Same prompt, both one shot on web-chat ui Qwen-3.8-27b produces consistent beautiful, visually stunning results. This is one of the best water surface simulations i have seen across every local models i have run and at the same time beating Gemini-3.7-flash result. And this is running on my 3090. And yes its Q5, not even f8.

Here's a comparison i never thought i would make Qwen-3.8-27b vs Gemini-3.7-flash Yes, you read it right! I don't know what to say, this is soo damn impressive. the difference is hugeee🤯 Take a look at this result 😍 (watch it in highest quality) Same prompt, both one shot on web-chat ui Qwen-3.8-27b produces consistent beautiful, visually stunning results. This is one of the best water surface simulations i have seen across every local models i have run and at the same time beating Gemini-3.7-flash result. And this is running on my 3090. And yes its Q5, not even f8.

AJ

302,905 görüntüleme • 2 ay önce

Do we need to build a Photoshop-like image rotation feature in ComfyUI? Tried to recreate PS’s image rotation feature inside ComfyUI. The tech itself isn’t anything special — it uses Qwen multi-angle, runs 32 angle variations of the image, then throws them onto a simple Canvas. The result obviously can’t compete with PS. Theirs clearly uses some kind of 3D technique under the hood. On top of that, a single run on my local 5090 takes 320 seconds — which is nowhere near an acceptable turnaround time. Just playing around with it for now. We’ll see if we can improve it later by switching to a 3D-based approach.

Do we need to build a Photoshop-like image rotation feature in ComfyUI? Tried to recreate PS’s image rotation feature inside ComfyUI. The tech itself isn’t anything special — it uses Qwen multi-angle, runs 32 angle variations of the image, then throws them onto a simple Canvas. The result obviously can’t compete with PS. Theirs clearly uses some kind of 3D technique under the hood. On top of that, a single run on my local 5090 takes 320 seconds — which is nowhere near an acceptable turnaround time. Just playing around with it for now. We’ll see if we can improve it later by switching to a 3D-based approach.

jtydhr88

14,790 görüntüleme • 7 ay önce

Parameter scaling just broke in front of me. I gave 90 math problems as images to 10 local vision models, 3 runs each, only consistent answers across all 3 runs counted. Two findings: > Gemma 4 was the most consistent family, 31B took the crown at 89.6%. > But Qwen 3.5 4B came 2 answers behind it. A 4B. Ranked 2nd out of 10. Vision-math isn't one skill, it's two: reading the image, then solving it. The real lesson for anyone running local: small ≠ weak. If you're building agentic workflows, finding the right model for each task matters more than reaching for the biggest one. In this test, the 4B ran far faster because of its size, scored higher, and left VRAM free for the rest of your stack. Full results: 🥇 Gemma 4 31B — 242/270 (89.6%) 🥈 Qwen 3.5 4B — 240/270 (88.9%) 🥉 Gemma 4 E4B — 222/270 (82.2%) 🥉 Qwen 3.6 27B — 222/270 (82.2%) 5. Gemma 4 26B A4B — 216/270 (80.0%) 6. Qwen 3.5 2B — 201/270 (74.4%) 7. Gemma 4 E2B — 192/270 (71.1%) 8. Qwen 3.6 35B A3B — 192/270 (71.1%) 9. Qwen 3.5 9B — 168/270 (62.2%) 10. Qwen 3.5 0.8B — 45/270 (16.7%) All GGUF + mmproj, Unsloth's Q6_K_XL quant.

Parameter scaling just broke in front of me. I gave 90 math problems as images to 10 local vision models, 3 runs each, only consistent answers across all 3 runs counted. Two findings: > Gemma 4 was the most consistent family, 31B took the crown at 89.6%. > But Qwen 3.5 4B came 2 answers behind it. A 4B. Ranked 2nd out of 10. Vision-math isn't one skill, it's two: reading the image, then solving it. The real lesson for anyone running local: small ≠ weak. If you're building agentic workflows, finding the right model for each task matters more than reaching for the biggest one. In this test, the 4B ran far faster because of its size, scored higher, and left VRAM free for the rest of your stack. Full results: 🥇 Gemma 4 31B — 242/270 (89.6%) 🥈 Qwen 3.5 4B — 240/270 (88.9%) 🥉 Gemma 4 E4B — 222/270 (82.2%) 🥉 Qwen 3.6 27B — 222/270 (82.2%) 5. Gemma 4 26B A4B — 216/270 (80.0%) 6. Qwen 3.5 2B — 201/270 (74.4%) 7. Gemma 4 E2B — 192/270 (71.1%) 8. Qwen 3.6 35B A3B — 192/270 (71.1%) 9. Qwen 3.5 9B — 168/270 (62.2%) 10. Qwen 3.5 0.8B — 45/270 (16.7%) All GGUF + mmproj, Unsloth's Q6_K_XL quant.

stevibe

22,882 görüntüleme • 5 ay önce

🚨Claude Opus 5.5 VS GPT 6.1 Sol. i gave both models the same simple prompt: create an animated website from a single image. Opus took around 1hr 20mins, while GPT-6.1 finished in just 15mins. both came up with really nice concepts, but Opus clearly spent more time on the details and delivered the better overall quality. GPT-6.1 wins on speed. Opus wins on execution. you can check both websites in the comments and decide for yourself.

🚨Claude Opus 5.5 VS GPT 6.1 Sol. i gave both models the same simple prompt: create an animated website from a single image. Opus took around 1hr 20mins, while GPT-6.1 finished in just 15mins. both came up with really nice concepts, but Opus clearly spent more time on the details and delivered the better overall quality. GPT-6.1 wins on speed. Opus wins on execution. you can check both websites in the comments and decide for yourself.

m0h

23,104 görüntüleme • 8 gün önce

This is probably the most complex workflow I’ve ever built, only with open-source tools. It took my 4 days. It takes four inputs: author, title, and style; and generates a full visual animated story in one click in ComfyUI . I worked on it for four days. There are still some bugs, but here’s the first preview. Here’s a quick breakdown: - The four inputs are sent to LLMs with precise instructions to generate: first, prompts for images and image modifications; second, prompts for animations; third, prompts for generating music. - All voices are generated from the text and timed precisely, as they determine the length of each animation segment. - The first image and video are generated to serve as the title, but also as the guide for all other images created for the video. - Titles and subtitles are also added automatically in Comfy. - I also developed a lot of custom nodes for minor frame calculations, mostly to match audio and video. - The full system is a large loop that, for each line of text, generates an image and then a video from that image. The loop was the hardest part to build in this workflow, so it can process either a 20-second video or a 2-minute video with the same input. - There are multiple combinations of LLMs that try to understand the text in the best way to provide the best prompts for images and video. - The final video is assembled entirely within ComfyUI. - The music is generated based on the LLM output and matches the exact timing of the full animation. - Done! For reference, this workflow uses a lot of models and only works on an RTX 6000 Pro with plenty of RAM. My goal is not to replace humans, as I’ll try to explain later, this workflow is highly controlled and can be adapted or reworked at any point by real artists! My aim was to create a tool that can animate text in one go, allowing the AI some freedom while keeping a strict flow. I don’t know yet how I’ll share this workflow with people, I still need to polish it properly, but maybe through Patreon. Anyway, I hope you enjoy my research, and let’s always keep pushing further! :)

This is probably the most complex workflow I’ve ever built, only with open-source tools. It took my 4 days. It takes four inputs: author, title, and style; and generates a full visual animated story in one click in ComfyUI . I worked on it for four days. There are still some bugs, but here’s the first preview. Here’s a quick breakdown: - The four inputs are sent to LLMs with precise instructions to generate: first, prompts for images and image modifications; second, prompts for animations; third, prompts for generating music. - All voices are generated from the text and timed precisely, as they determine the length of each animation segment. - The first image and video are generated to serve as the title, but also as the guide for all other images created for the video. - Titles and subtitles are also added automatically in Comfy. - I also developed a lot of custom nodes for minor frame calculations, mostly to match audio and video. - The full system is a large loop that, for each line of text, generates an image and then a video from that image. The loop was the hardest part to build in this workflow, so it can process either a 20-second video or a 2-minute video with the same input. - There are multiple combinations of LLMs that try to understand the text in the best way to provide the best prompts for images and video. - The final video is assembled entirely within ComfyUI. - The music is generated based on the LLM output and matches the exact timing of the full animation. - Done! For reference, this workflow uses a lot of models and only works on an RTX 6000 Pro with plenty of RAM. My goal is not to replace humans, as I’ll try to explain later, this workflow is highly controlled and can be adapted or reworked at any point by real artists! My aim was to create a tool that can animate text in one go, allowing the AI some freedom while keeping a strict flow. I don’t know yet how I’ll share this workflow with people, I still need to polish it properly, but maybe through Patreon. Anyway, I hope you enjoy my research, and let’s always keep pushing further! :)

Lovis Odin

58,950 görüntüleme • 1 yıl önce

After analyzing 100+ OpenClaw mission controls, I cracked the code on building them. I built this one in one-prompt with Lovable. The full prompt for this is below. Give it a try in Lovable↓ I recorded an entire video on how to connect Lovable + OpenClaw which I'll include as well. PS. I'd usually gate-keep this but if you find it valuable - I'd just ask you show some love in the comments + follow.

After analyzing 100+ OpenClaw mission controls, I cracked the code on building them. I built this one in one-prompt with Lovable. The full prompt for this is below. Give it a try in Lovable↓ I recorded an entire video on how to connect Lovable + OpenClaw which I'll include as well. PS. I'd usually gate-keep this but if you find it valuable - I'd just ask you show some love in the comments + follow.

Jacob Klug

15,564 görüntüleme • 7 ay önce

in his interview, #LegendOfZangHai co-director cao yiwen said there's a scene in the drama in which #XiaoZhan left a deep impression on him. zang hai discovered that gao ming shifu who had been with him for 10 years had deceived him for 10 years. this scene could be interpreted as a fierce quarrel, but what the director wanted was for zang hai and gao ming to chat about daily life, saying light words, with eyes revealing silent intention, zang hai hoping that gao ming would tell him the truth and reveal the truth, and he was even willing to forgive. but gao ming avoided it, and zang hai was extremely disappointed. zang hai did not have a single line in the scene (at the end of the convo). "it is ten times more difficult to act with restraint than to act in flashy way. through his interpretation, xiao zhan showed people the psychological changes of zang hai at that time. it is easy to act in a melodramatic way, but it is difficult to act with restraint and control." this scene is so well-acted 😭♥️ you could see just how bitter and sad zang hai was but he had to keep his inner turmoil under control. xiao zhan's expression at the end of the convo was heartbreaking 💔😫

in his interview, #LegendOfZangHai co-director cao yiwen said there's a scene in the drama in which #XiaoZhan left a deep impression on him. zang hai discovered that gao ming shifu who had been with him for 10 years had deceived him for 10 years. this scene could be interpreted as a fierce quarrel, but what the director wanted was for zang hai and gao ming to chat about daily life, saying light words, with eyes revealing silent intention, zang hai hoping that gao ming would tell him the truth and reveal the truth, and he was even willing to forgive. but gao ming avoided it, and zang hai was extremely disappointed. zang hai did not have a single line in the scene (at the end of the convo). "it is ten times more difficult to act with restraint than to act in flashy way. through his interpretation, xiao zhan showed people the psychological changes of zang hai at that time. it is easy to act in a melodramatic way, but it is difficult to act with restraint and control." this scene is so well-acted 😭♥️ you could see just how bitter and sad zang hai was but he had to keep his inner turmoil under control. xiao zhan's expression at the end of the convo was heartbreaking 💔😫

ᴀɪʀɪ ₍ᐢ.ˬ.ᐢ₎ | 🛠️ 𝗱𝗲𝘅𝗶𝗮𝗻 𝘁𝗶𝗺𝗲!

14,762 görüntüleme • 1 yıl önce

hey here is the final result of octopus invaders on nvidia's flagship at full precision. nemotron super 120B on 2x H200 NVL. BF16 unquantized. 287GB of VRAM. hermes agent as the harness. 60 tok/s. first try it autonomously coded for 6 minutes straight. created 11 files. correct project structure. correct load order. started the server. i opened the browser and the result was a blank screen. i did not give up. second try i gave it a precise list of bugs and things to fix. it went back in for another 3 minutes. patched the code. served it again. still blank. so i did what any sane person would do. third try i just said the screen is blank, test it and fix it yourself. and this is where nemotron showed what it actually is. it became a debugger. you can see it in the video. realtime CSS test squares, red screen flashes, hermes agent browser tools, inspecting its own output. it built the parallax background with planets and comets. it rendered a rocket ship that tracks your mouse with fire and bullet physics. the aesthetic is real. but no enemies spawn. no collision. not playable. what surprised me is qwen 27B one shotted this exact game on a single RTX 3090 at Q4 quant. and here is nvidia's flagship at full precision on enterprise hardware needing 3 tries and still not getting there. that makes my hope high for the undisputed qwen 122B which is about to face the same test next. same hardware. same prompt and same harness. lets see if it one shots or not. full session in the video. no cuts. 5x speed.

hey here is the final result of octopus invaders on nvidia's flagship at full precision. nemotron super 120B on 2x H200 NVL. BF16 unquantized. 287GB of VRAM. hermes agent as the harness. 60 tok/s. first try it autonomously coded for 6 minutes straight. created 11 files. correct project structure. correct load order. started the server. i opened the browser and the result was a blank screen. i did not give up. second try i gave it a precise list of bugs and things to fix. it went back in for another 3 minutes. patched the code. served it again. still blank. so i did what any sane person would do. third try i just said the screen is blank, test it and fix it yourself. and this is where nemotron showed what it actually is. it became a debugger. you can see it in the video. realtime CSS test squares, red screen flashes, hermes agent browser tools, inspecting its own output. it built the parallax background with planets and comets. it rendered a rocket ship that tracks your mouse with fire and bullet physics. the aesthetic is real. but no enemies spawn. no collision. not playable. what surprised me is qwen 27B one shotted this exact game on a single RTX 3090 at Q4 quant. and here is nvidia's flagship at full precision on enterprise hardware needing 3 tries and still not getting there. that makes my hope high for the undisputed qwen 122B which is about to face the same test next. same hardware. same prompt and same harness. lets see if it one shots or not. full session in the video. no cuts. 5x speed.

Sudo su

11,011 görüntüleme • 6 ay önce

People think Matt faked it when he said one prompt built a whole cod game. So i read the prompt properly and ran it myself. Spent over 30% of my weekly tokens on the 20x plan for 12 hours. It's real. but here's the part nobody's talking about. The prompt says don't stop until you beat call of duty in a blind side by side. It never beat it. It kept making the scores, and judges the game every single round. it failed its own test the entire time. That's why the game got better and better . An impossible high bar means it never stops improving.

People think Matt faked it when he said one prompt built a whole cod game. So i read the prompt properly and ran it myself. Spent over 30% of my weekly tokens on the 20x plan for 12 hours. It's real. but here's the part nobody's talking about. The prompt says don't stop until you beat call of duty in a blind side by side. It never beat it. It kept making the scores, and judges the game every single round. it failed its own test the entire time. That's why the game got better and better . An impossible high bar means it never stops improving.

Ziwen

86,753 görüntüleme • 2 ay önce

Very quick comparison between Ornith 1.5 35B MoE and Qwen 3.8 27B Dense. 📣 Clearly it's not a fair one, but let's in any case see how it went. 397B download in progress! In the Videos below: - Brick SA -> Qwen 3.8 - Lego Streets -> Ornith 1.5 Context: - Pi agent used in both cases, prompt below - M3 Ultra 512GB - Ornith-1.5-35B-A3B-oQ8e MTP hosted on oMLX - mlx-community/Qwen3.8-27B-8bit DFlash 2 hosted on mlx-dspark - Ornith has been incredibly fast with speed from 75 t/s to 45 t/s (160K+ context) - Qwen3.8 suffered context much more reaching 5 t/s above 160K context, but it can be due to engine tested still work in progress - Prompt: using threejs, and cdn, create a lego like game that's inspired by gta san andreas, with beautiful aesthetics and graphics and ability to steal cars. it should be 3d and have a nice and large map and areas. the graphics should be decent and nice. it should feature iconic things from gta san andreas Notes: - Qwen 3.8 result is 0-shot, while Ornith 1.5 required 6 iterations - Ornith 1.5 is not at the same level of autonomy as Qwen 3.8 27B honestly. To try getting the same results I'm constantly nudging, steering and 🤬 at it. - Qwen 3.8 took 6 hours to complete, but it was more a problem of timeout of Chrome headless used for testing, real prefill/decode TBD. I'll test again now with oMLX - Pi agent has a nearly perfect Cache Hit ratio that for local models is MEGA important!

Very quick comparison between Ornith 1.5 35B MoE and Qwen 3.8 27B Dense. 📣 Clearly it's not a fair one, but let's in any case see how it went. 397B download in progress! In the Videos below: - Brick SA -> Qwen 3.8 - Lego Streets -> Ornith 1.5 Context: - Pi agent used in both cases, prompt below - M3 Ultra 512GB - Ornith-1.5-35B-A3B-oQ8e MTP hosted on oMLX - mlx-community/Qwen3.8-27B-8bit DFlash 2 hosted on mlx-dspark - Ornith has been incredibly fast with speed from 75 t/s to 45 t/s (160K+ context) - Qwen3.8 suffered context much more reaching 5 t/s above 160K context, but it can be due to engine tested still work in progress - Prompt: using threejs, and cdn, create a lego like game that's inspired by gta san andreas, with beautiful aesthetics and graphics and ability to steal cars. it should be 3d and have a nice and large map and areas. the graphics should be decent and nice. it should feature iconic things from gta san andreas Notes: - Qwen 3.8 result is 0-shot, while Ornith 1.5 required 6 iterations - Ornith 1.5 is not at the same level of autonomy as Qwen 3.8 27B honestly. To try getting the same results I'm constantly nudging, steering and 🤬 at it. - Qwen 3.8 took 6 hours to complete, but it was more a problem of timeout of Chrome headless used for testing, real prefill/decode TBD. I'll test again now with oMLX - Pi agent has a nearly perfect Cache Hit ratio that for local models is MEGA important!

Ivan Fioravanti ᯅ

12,150 görüntüleme • 1 ay önce