monkeys dot im (@monkeys_im) header
monkeys dot im (@monkeys_im) profile picture

Important work deserves the best AI. Slides, images, video and research, ranked by public evals with real cost shown. Follow for each new model and agent eval.

Joined November 2026
Photo by monkeys dot im (@monkeys_im)Photo by monkeys dot im (@monkeys_im)Photo by monkeys dot im (@monkeys_im)Photo by monkeys dot im (@monkeys_im)Photo by monkeys dot im (@monkeys_im)Photo by monkeys dot im (@monkeys_im)Photo by monkeys dot im (@monkeys_im)Photo by monkeys dot im (@monkeys_im)Photo by monkeys dot im (@monkeys_im)Photo by monkeys dot im (@monkeys_im)
GPT-6.2 Sol just came out. Can it walk? Same prompt for all four: a person walking across the screen in a natural, looping walk cycle. GPT-6.2 Sol vs GPT-6.1 Sol, Claude Opus 5.5 and Gemini 4 Carbon. Which walk looks the most natural? Try your own prompt on these models.

5
4
11,359
Tap the video card above for the eval page. The full prompt, every output, token count and cost are all there.
90
GPT-6.2 Sol just came out. Can it walk? GPT-6.2 Sol vs GPT-6.1 Sol, Claude Opus 5.5 and Gemini 4 Carbon. Same prompt: a person walking in a natural, looping walk cycle. It's 1 of 6 real jobs in this eval. The full prompt, every output, token count and cost are public 👇

5
3
7,694
Tap the video card above for the eval page. The full prompt, every output, token count and cost are all there.
78
GPT-6.2 Astra just came out. Time for the classic pelican test. Same prompt for GPT-6.2 Astra, GPT-6.2 Sol, Claude Opus 5.5 and Claude Fable 5.2. Full prompt, every output, token count and cost 👇 It's 1 of 6 real jobs in this eval.

5
19
19,870
Tap the video card above for the eval page. The full prompt, every output, token count and cost are all there.
367
GPT-6.2 Sol just came out. Can it walk? GPT-6.2 Sol vs GPT-6.1 Sol, Claude Opus 5.5 and Gemini 4 Carbon. Same prompt: a person walking in a natural, looping walk cycle. Tap the card for the full prompt, every output, token count and cost. It's 1 of 6 real jobs in this eval.

3
10
10,694
Tap the video card above for the eval page. The full prompt, every output, token count and cost are all there.
122
GPT-6.2 Astra just came out. Time for the classic pelican test. Same prompt for GPT-6.2 Astra, GPT-6.2 Sol, Claude Opus 5.5 and Claude Fable 5.2. Tap the card for the full prompt, every output, token count and cost. It's 1 of 6 real jobs in this eval.

1
2,360
Tap the video card above for the eval page. The full prompt, every output, token count and cost are all there.
23
You ask Grok things on X every day. How good is it, really? Grok cost 3 to 8x more. DeepSeek took up to 2.5x longer. Tap the card for the full prompts, every output, token counts and costs. 3 of 6 jobs in this eval.

2
1
1,476
Tap the video card above for the eval page. The full prompts, every output, token counts and costs are all there.
6
You ask Grok things on X every day. How good is it, really? Grok cost 3 to 8x more. DeepSeek took up to 2.5x longer. Tap the card for the full prompts, every output, token counts and costs. 3 of 6 jobs in this eval.
🤖 Made with AI

2
9
13,842
Tap the video card above for the eval page. The full prompts, every output, token counts and costs are all there.
71
You ask Grok things on X every day. How good is it, really? Grok cost 3 to 8x more. DeepSeek took up to 2.5x longer. Full prompts, every output, token counts and costs are below. 3 of 6 jobs in this eval 👇
3
1
1
979
Follow @monkeys_im so you don't miss the next match-up.
30
GPT-6.2 Sol vs GPT-6.1 Sol, Gemini 4 Carbon and Claude Opus 5.5. Same prompt: a miniature 3D fishing harbor in a sunshower. Costs ran from $0.28 to $7.72. Orbit every harbor yourself, then run your own prompt on these models.
🤖 Made with AI

2
1
11
16,815
GPT-6.1 Terra vs Qwen4 Max, Claude Sonnet 5.5 and Grok 4.7. Can they walk? Same prompt: a person walking in a natural, looping walk cycle. Costs ran from $0.055 to $2.81. It's 1 of 6 real jobs in this eval. The full prompt, every output, token count and cost are public 👇
🤖 Made with AI

1
3
5,349
GPT-6.2 Astra just came out. Did Astra finally learn to walk? Same walk-cycle prompt for all four. GPT-6.2 Astra vs GPT-6 Astra, Claude Opus 5.5 and Claude Fable 5.2. Full prompt, every output, token count and cost 👇 It's 1 of 6 real jobs in this eval.
🤖 Made with AI

2
1,664
GPT-6.2 Sol just came out. Can it walk? GPT-6.2 Sol vs GPT-6.1 Sol, Claude Opus 5.5 and Gemini 4 Carbon. Same prompt: a person walking in a natural, looping walk cycle. Tap the card for the full prompt, every output, token count and cost. It's 1 of 6 real jobs in this eval.
🤖 Made with AI

7
3
2
59
96,026
GPT-6.2 Astra just came out. Time for the classic pelican test. Same prompt for GPT-6.2 Astra, GPT-6.2 Sol, Claude Opus 5.5 and Claude Fable 5.2. Tap the card for the full prompt, every output, token count and cost. It's 1 of 6 real jobs in this eval.
🤖 Made with AI

6
3
2
53
63,015
GPT-6.2 Astra just came out. Time for the classic pelican test. Same prompt for GPT-6.2 Astra, GPT-6.2 Sol, Claude Opus 5.5 and Claude Fable 5.2. Full prompt, every output, token count and cost 👇 It's 1 of 6 real jobs in this eval.
4
1
1
693
New models get the same jobs as they come out. Follow @monkeys_im so you don't miss the next one.
53
GPT-6.2 Astra just came out. Did Astra finally learn to walk? Same walk-cycle prompt for all four. GPT-6.2 Astra vs GPT-6 Astra, Claude Opus 5.5 and Claude Fable 5.2. Full prompt, every output, token count and cost 👇 It's 1 of 6 real jobs in this eval.
3
1
1
1,163
New models get the same jobs as they come out. Follow @monkeys_im so you don't miss the next one.
48