Timur Ludanov

Proud to share what we’ve been building at EidoStack 🚀 Read the launch story and follow along as we make AI model evaluation more practical for real-world products.

EidoStack update: • Delete individual chat messages • 2 new context strategies: None — current prompt only System Prompt Only — system + prompt (no history) eidostack.com/blog/more-co...

🚀 Gemini 4 Argon is now available in EidoStack! Explore Google's latest AI model in our Playground. Experiment with prompts, explore responses, and track token usage and API costs using your own Google API key. eidostack.com/blog/gemini-...

Claude Haiku 5.5 is now supported in EidoStack! 🎉 Try it in the Playground with your Anthropic API key. Experiment with prompts and track token usage. Read more 👇 eidostack.com/blog/claude-...

GPT-6 Sol and GPT-6 Luna are now available in EidoStack. You can now put both models side by side, test the same prompt, and compare response quality, token usage, and cost before using them in production. Read more: eidostack.com/blog/gpt-6-s... #AI #LLM #OpenAI

Claude Opus 5.5 is now available in EidoStack. Test it on your real workloads, inspect token usage and cost, and compare the results with GPT, Gemini, or other Claude models. eidostack.com/blog/claude-...

GPT-6 Astra is now available in EidoStack. It’s now available alongside the other supported models, so you can evaluate Astra in the same workspace and see where its extra capability actually makes a difference. Read more: eidostack.com/blog/gpt-6-a...

Gemini 3.8 Flash is now available in EidoStack. A new model is only interesting if it performs well on your actual workload. You can now evaluate Gemini 3.8 Flash in EidoStack and see how it behaves against the models you already use. Read more: eidostack.com/blog/gemini-...

Claude Fable 5.1 has landed in EidoStack. You can now test it, compare it with other models side by side, and see how it performs on your own prompts. More details: eidostack.com/blog/claude-...

Noticed today that Codex seems to have brought back the 5-hour usage window alongside the weekly limit. For a while, I was only seeing the weekly limit, which was much more convenient. The 5-hour window makes longer agent sessions harder to plan. I definitely preferred the weekly-only setup.

Been building EidoStack for a while, and now it finally has its own Bluesky account. I’ll use it to share product updates, practical AI model comparisons, and experiments around real development workflows. Follow @eidostack.com if that sounds useful.

Hello Bluesky — we’re EidoStack. A workspace where you can chat with AI models, compare their responses side by side, and evaluate quality, token usage, and cost for real applications. We’ll share product updates, experiments, and practical model comparisons. eidostack.com

Google just announced Gemini 3.7 Flash - I got the news from Google AI Studio. Most interesting to me: better software engineering and agentic workflows, fewer failed agent loops, and stronger design-to-code. Intro pricing: $0.75 / 1M input, $3.75 / 1M output until Dec 31, 2026. I want to test it

Starting to use Bluesky. Mostly interested in software engineering, AI, LLMs, SaaS, and building products with AI. Looking forward to sharing ideas and learning from others.