←
Tony8,312 posts
Tony avatar

Tony

@EnvolDev

Racing top AI models: same prompt, real cost & time Building SaaS apps & tools with AI First on everything new, and only what matters laeli.app & clipfame.app

◷ Joined August 2026▧ 275 Photos and videos
Tony avatar
Tony@EnvolDev·23m

Hey @grok, renting a home or buying one: which will a normal 30-year-old regret more by 2035? Pick one and explain why.

Tony avatar
Tony@EnvolDev·6h

GPT-6.2 Sol built my 4K flamingo film in 32 minutes. Now I definitely want to try that same prompt on Ultrafast. 👀

If you have Pro 200/500, I do recommend trying out 6.2 Sol ultrafast, it's quite the blast to use and a completely different feel. I'm excited to keep pushing in bringing these speeds to more users over time.

Tony avatar
⟳ Tony retweeted
Tony@EnvolDev·16h

GPT-6.2 Astra vs GPT-6.2 Sol vs Fable 5.2 vs Opus 5.5 🛩️ GPT-6.2 Astra is just f*cking insane. 🤯 OpenAI is so f*cking back. Over 21x cheaper and almost 4x faster than Fable 5.2. And the detail on this plane is unreal. 😱 Same Alpine Gold prompt: GPT-6.2 Astra: $9.62 · 46m 14s GPT-6.2 Sol: $2.07 · 30m 47s Fable 5.2: $209.50 · 3h 02m 54s Opus 5.5: $31.85 · 2h 45m 42s And GPT-6.2 Sol did it for $2.07. Over 100x cheaper than Fable 5.2. 💀 Is Anthropic in trouble? 👀

0:25
Tony@EnvolDevNov 10

Tony @EnvolDev Nov 10 Article The AI Masterclass: Why the Best Model Is Rarely the Right One How to build real things with AI, and which model to use for each job: from your first prompt to your first app to a team of agents. I don't write code. I build by directing AI agents: 3D films, SaaS

Tony avatar
⟳ Tony retweeted
Tony@EnvolDev·17h

GPT-6.2 Sol vs Opus 5.5 🛩️ GPT-6.2 Sol is just insane. 🤯 It's over 15x cheaper and over 5x faster than Opus 5.5. And the plane detail is out of this world. 😱 Same Alpine Gold prompt: GPT-6.2 Sol: $2.07 · 30m 47s Opus 5.5: $31.85 · 2h 45m 42s $2 for this. 💀 Honestly, this one is really hard to judge. GPT-6.2 Sol just made Opus 5.5 look overpriced. Is OpenAI back? 👀

0:25
Tony@EnvolDevNov 10

Tony @EnvolDev Nov 10 Article The AI Masterclass: Why the Best Model Is Rarely the Right One How to build real things with AI, and which model to use for each job: from your first prompt to your first app to a team of agents. I don't write code. I build by directing AI agents: 3D films, SaaS

Tony avatar
⟳ Tony retweeted
Tony@EnvolDev·18h

GPT-6.2 Astra vs Fable 5.2 🧊 GPT-6.2 Astra is almost 10x cheaper and almost 3x faster than Fable 5.2. 🤯 But the quality is nowhere near Fable 5.2. 😬 Mixed feelings. Same Cathedral of Ice prompt: Fable 5.2: $81.10 · 1h 56m 52s GPT-6.2 Astra: $8.20 · 40m 34s For coding and debugging, GPT-6.2 Astra is going to be a game changer. For UI design and visuals, Fable 5.2 is still far ahead. Cheap and fast, or the best quality? 👀

0:25
Tony@EnvolDevNov 10

Tony @EnvolDev Nov 10 Article The AI Masterclass: Why the Best Model Is Rarely the Right One How to build real things with AI, and which model to use for each job: from your first prompt to your first app to a team of agents. I don't write code. I build by directing AI agents: 3D films, SaaS

Tony avatar
Tony@EnvolDev·11h

OpenAI just shocked the world... I waited 75 days for a new Astra. It took 46 minutes to change my mind. 🤯 OpenAI had shelved GPT-6.1 Astra. Anthropic dropped Fable 5.2. Google dropped Gemini 4 Carbon and Hydrogen. I honestly thought OpenAI was falling behind. Then GPT-6.2 Astra and Sol both dropped today. If you saw my Alpine Gold benchmark, look at the detail on Astra’s plane. It built that in 46 minutes for an estimated $9.62. Sol built its version for $2.07. Then I gave all three models the same 4K flamingo film to build. Here’s how long each took and what each run cost at API rates. Fable 5.2 — 2h 07m · $82.18 GPT-6.2 Astra — 46m · $10.02 GPT-6.2 Sol — 32m · $1.98 Astra finished almost 3× as fast for about an eighth of the cost. And Sol made a whole 4K film for under $2. I’m keeping Opus 5.5 for design and UI. But after seeing that plane, I’m giving ChatGPT the same design work to see which one I’d actually use. For coding and debugging, I was already using ChatGPT. At this speed and price, I’m all in. Is it AGI? I don’t know. But after what I saw today, it feels pretty f*cking close. 👀

Tony avatar
Tony@EnvolDev·19h

GPT-6.2 Sol taking #1 ahead of Fable 5.2… If this is right, this is crazy. 🤯 Let’s see what my benchmarks show. 👀

Post attachment
Tony avatar
Tony@EnvolDev·20h

Already cooking some benchmarks here!!! GPT-6.2-Astra and GPT-6.2-Sol 💪👀

Post attachment
Tony avatar
Tony@EnvolDev·20h

GPT 6.2 Sol at the same time! 😳

Tony avatar
Tony@EnvolDev·20h

OMG! 😱 GPT 6.2 Astra is released !!! LFG!!!!

Post attachment
Tony avatar
Tony@EnvolDev·21h

Fable 5.2 vs Opus 5.5 🦩 Very close fight on this one. 😳 Same Rosewater Flight prompt: Fable 5.2: $82.18 · 2h 07m 08s Opus 5.5: $57.57 · 4h 15m 23s Opus 5.5 took twice as long. 🤯 Fable 5.2 cost $24.61 more. 😱 Is Fable 5.2 worth the extra money? 👀

0:25
Tony@EnvolDevNov 10

Tony @EnvolDev Nov 10 Article The AI Masterclass: Why the Best Model Is Rarely the Right One How to build real things with AI, and which model to use for each job: from your first prompt to your first app to a team of agents. I don't write code. I build by directing AI agents: 3D films, SaaS

Tony avatar
Tony@EnvolDev·22h

GPT-6.2 Astra and Sol could finally drop today. 🤯 After all this waiting, “slightly better” won’t be enough. OpenAI, give us something that makes Fable 5.2 look old. 👀

Tony avatar
Tony@EnvolDev·23h

Yesterday, I got accepted into the Original Content Rewards Program after just 75 days on X. Today, I passed 3,000 followers! 🥳 Thank you for every follow, comment, and like. You make all the hard work worth it. More benchmarks, more builds, and more articles like the AI Masterclass are coming. I’m going to keep working harder to bring you as much value as possible. Let’s keep growing together! 🙏

Post attachment
Tony avatar
Tony@EnvolDev·Nov 17

Hey @grok, by 2030 AI agents will be shopping too. What can someone sell to AI agents, not humans, that pays well? Name it.

Tony avatar
⟳ Tony retweeted
Tony@EnvolDev·Nov 16

Qwen 4 Max vs Fable 5.2 vs Gemini 4 Argon vs GPT-6 Astra 🛩️ Alibaba's top model vs three American frontier models. 🇨🇳🇺🇸 Same Alpine Gold prompt: Qwen 4 Max: $12.03 · 4h 56m 36s Fable 5.2: $209.50 · 3h 02m 54s Gemini 4 Argon: $11.61 · 2h 01m 35s GPT-6 Astra: $17.82 · 51m 10s Qwen 4 Max worked for almost 5 hours. 🤯 Longer than Fable 5.2. Almost 6x longer than GPT-6 Astra. Now look at what it made. 😬 Almost 5 hours for that? 😱 Qwen 4 Max is nowhere near the frontier. Not even close. Is Qwen 4 Max overhyped? 👀

0:25
Tony@EnvolDevNov 10

Tony @EnvolDev Nov 10 Article The AI Masterclass: Why the Best Model Is Rarely the Right One How to build real things with AI, and which model to use for each job: from your first prompt to your first app to a team of agents. I don't write code. I build by directing AI agents: 3D films, SaaS