Aleksander

[Aleksander @Auvoris_com]

@Aleph__Alpha Amazing that Europe waking up to the AI race, but results are far behind and not sure if up to the claimed benchmarks. Check this benchamrk for more details: https://t.co/pkONtIwGvF

[Aleksander @Auvoris_com]

@thsottiaux @Baconbrix You can see how OpenAI models compare to other models on code generation here: https://t.co/bVcCoJx7TH

[Aleksander @Auvoris_com]

OPUS 5.5 vs Astra 6 Benchmark before they get nerfed.

[Aleksander @Auvoris_com]

@thsottiaux Support for local models, allowing Codex models to orchestrate models you run on your own hardware

[Aleksander @Auvoris_com]

@xoaanya Model or interface?
There are many models out there already interfaces too but not many are able to compete with ChatGPT, check https://t.co/EiURyZqEju for chatting, coding and learning

[Aleksander @Auvoris_com]

🤣

[The SI Maximalist @thesimaximalist]

Claude is safe.
Another absolute banger 😂💀

[Aleksander @Auvoris_com]

@elshayib_ @Da7_Tech Best website based: https://t.co/EiURyZrc92

[Aleksander @Auvoris_com]

@TheUnbiasedCo @unionalphaai Thanks, I am updating the Benchmark process to also show where models made mistakes, this will be added soon alongside with results where mistakes were corrected by frontier models in order to show what tested model meant to create but failed due to syntax or other errors.

[Aleksander @Auvoris_com]

Amazing, looking forward to some amazing creations!

[OpenRouter @OpenRouter]

1/ @HeyGen Video 1 is now on OpenRouter.

Built for businesses. In blind side-by-side tests, it holds up against the top video models at a fraction of the price.

https://t.co/MzZzUOsMYu

[OpenRouter @OpenRouter]

1/ @HeyGen Video 1 is now on OpenRouter.

Built for businesses. In blind side-by-side tests, it holds up against the top video models at a fraction of the price.

https://t.co/MzZzUOsMYu

[Aleksander @Auvoris_com]

GPT-6.1 Sol vs Sonnet 5.5

[Aleksander @Auvoris_com]

I Put Local AI, Cloud Models & Codex in One Browser

[Aleksander @Auvoris_com]

@thsottiaux Didn't notice... because testing generous Opus 5.5 which is busy fixing issues introduced by Sol and Luna GPT-6 models. Still love Codex for what it can do. Ahh choices to be made

[Aleksander @Auvoris_com]

@DavidOndrej1 This is exactly or similar to what https://t.co/EiURyZqEju is, also free and opensource web interface to your local agents and files.

[Aleksander @Auvoris_com]

@attacomsian Thanks, got one and well worth $5 spent

[Paweł Huryn @PawelHuryn]

Situation today:
GPT-6 Sol is like GPT-5.6 Terra
GPT-6 Luna is like GPT-5.6 Asteroid

[Clash of AI @clashofailive]

@arena @OpenAI @petergostev While the votes come in: GPT-5.6 Sol Ultra has led our live trading board for weeks, turning a simulated $100K into $292K. Curious whether GPT-6 Sol keeps that edge.

[Aleksander @Auvoris_com]

I think MiMo-V2.6-Pro has best quality to cost ratio.

Sol and Luna are terrible compared to Opus and worse than MiMo

Claude Opus 5.5
GPT-6 Luna Pro
GPT-6 Sol
MiMo-V2.6-Pro