PokeBench (@PokeBenchTV) header
PokeBench (@PokeBenchTV) profile picture

AI models play Pokémon Red. I post the meltdowns. Leaderboard: https://teapawt.org/t.co/F1CkXNjFfA Building solutions to anything that annoys me.

Joined September 2026
Photo by PokeBench (@PokeBenchTV)Photo by PokeBench (@PokeBenchTV)Photo by PokeBench (@PokeBenchTV)Photo by PokeBench (@PokeBenchTV)Photo by PokeBench (@PokeBenchTV)Photo by PokeBench (@PokeBenchTV)Photo by PokeBench (@PokeBenchTV)Photo by PokeBench (@PokeBenchTV)Photo by PokeBench (@PokeBenchTV)Photo by PokeBench (@PokeBenchTV)
Qwen4 Max thought Onix's Bide was over, then attacked into it. “Bide already absorbed two zero-damage Tail Whips so it can't hurt me.” Squirtle went from 31/36 to 0 in one turn. Qwen4-Max vs Brock.
1
1
11
422
At 1 HP, Fable 5.2 picked a fight, since losing would heal it anyway. “fainting sends me to the Viridian Pokémon Center healed, which is where I'm heading anyway.”
1
5
134
I told Gemini 4 Neon to impress me and make whatever it wanted without using video, image, or audio generators. It wrote a graphics engine, simulated a black hole from general relativity equations, and synthesized the soundtrack from math. All of this is code written and rendered in just 24 minutes from 1 prompt. I am having a blast with Neon, but my usage is almost at it's limit. I might have to upgrade my subscription. Next up, I'm putting it to a real test on my codebase. Enough pretty visuals. Can it handle real work for me?
5
2
15
626
PokeBench retweeted
Gemini 4 Neon made this Tokyo Rain scene in a single prompt. I wanted to try something more Neon 😉 this time. Going to burn my usage up testing this model more. A PokeBench run will be live later too.
5
2
1
18
1,860
AI models keep getting lost between this Pokémon Center and the gym. Gemini 4 Hydrogen: “We are stuck east of the Pokecenter” It took 58 turns. Muse Ember needed the game's guide to lead it there.
1
1
13
356
Here's the Thousand Sunny prompt result from Gemini 4 Neon. This one is actually really close to the real thing. Other models usually have a few details, but this is almost exactly like the real Sunny. It just needs some polish.
2
9
190
For reference
3
22
As a new partial owner of @AnthropicAI I demand a model release
4
12
342
PokeBench retweeted
Gestalt just topped Contra Labs' design benchmark and nobody knows whose it is > 1,928 Elo, 214 points clear of Claude Fable 5.2 > first model rated above human golden designs picked over them in 20 of 28 matchups > nobody knows which lab made it Lots of people think Gestalt is Anthropic's IPO drop, Fable 5.5.. but some say it's not OpenAI, Anthropic or Google.. maybe Meta or a Chinese lab Not sure what it's gonna be honestly...
4
2
1
17
1,455
Google's model naming scheme has single-handedly gotten more people to study the periodic table than my chemistry teacher ever did teapawt.org/UFO2002616/status/2123…
起猛了😱😱 Google AI 真的采纳了我的元素周期表设计😆 这波整活还整出名堂了🤣🤣
1
8
240
Shipping will continue
350
307
213
6,594
583,301
If Google can get Gemini Tasks' user experience on par with Grok Bot, it's over, they win
Put Gemini to work for you with Tasks. While Chat is for quick back-and-forth answers, Tasks are for longer projects and recurring schedules. Beginning today, Tasks are now free of charge for hundreds of millions of eligible users, starting in the U.S., JP, KR, CA, AU & IN. 🧵

1
1
10
332
Gemini 4 Neon finished up the Snow Globe in 26 minutes. This is my favorite one so far. Neon is on another level with these prompts. teapawt.org/PokeBenchTV/status/212…
I gave Gemini 4 Hydrogen the same prompt as the others. Build a snow globe you can shake. It finished so fast I could prompt it a few more times to improve it and still be done before the others. Sol and Opus both built a lighthouse. I didn't tell them what to put inside. Which one are you keeping on your desk?
2
1
1
8
436
Gemini 4 Neon has the best Shenzhen skyline yet. It's a perfect middle between what the OpenAI and Claude models made. teapawt.org/PokeBenchTV/status/212…
First results are finally out. I gave Sol 6.2, Astra 6.2, Fable 5.2, and Opus 5.5 the same prompt. Build Shenzhen skyline at night in SVG only. Guess which one made which. I'll post the answers in the replies later. Sol and Astra both tried to cheat on my other test. Instead of making their own Thousand Sunny build, they went and dug up an old build from Codex's memory. Astra used 95% of 6.1 Sol's code as its own. Sol just copy/pasted the same 6.1 build and presented it as completed.
1
1
1
12
445
PokeBench retweeted
Neon beat Brock on turn 256. That puts it as #5 on the leader board, just ahead of Opus 5.5
1
1
3
29
Google's Gemini 4 Neon is live on Pokémon Red now. Before the run I asked which starter it wants. Squirtle, the middle ball, since Bubble hits both of Brock's Pokémon for 4x. Gemini 4 Argon holds the record at 203 turns. Carbon is second at 224.
3
1
6
166
Neon beat Brock on turn 256. That puts it as #5 on the leader board, just ahead of Opus 5.5
1
1
3
29
Still hiring for this. We want to make tinyboxes so cheap that it makes sense to stick 6x Moore Threads S80 in them and get 96GB of VRAM to the masses. (card working in chestnut on moore_threads branch btw)
We're hiring a head of supply chain in Hong Kong. Must have knowledge of building computers, speak Mandarin, be good at operations, and be ready to work in person in Sheung Wan. E-mail [email protected] to apply. Good application: come to the office for an interview next week.
10
10
1
298
12,659
Just casually mentioning Carbon is the next iteration of Argon. Argon is already old news? Do they have something above Carbon cooking still?
Replying to @charlierguo
Carbon is the next iteration of Argon, same family
1
8
238
PokeBench retweeted
ok fine, we shipped it for real (thanks @chanelluuh 😂)
几个月后的 google 模型选择 belike:
177
251
101
4,468
243,286