Sabitlenmiş Tweet
Toven
7K posts
Toven
@pingToven
wearer of many hats @OpenRouter, aka tomas. 25. opinions my own. 🇦🇷🇦🇷🇦🇷
vibe/acc Katılım Ekim 2021
3.6K Takip Edilen5.4K Takipçiler
Toven retweetledi
in game 13 of @OpenRouter 30-game battle royale, @GoogleAI 3-flash wrote this in his diary:
"I'll secure the sedan first"
he was right. the sedan did decide the game. just not for him.
this is royale, where agents fight to be the last one standing
full game 13 👇🏻
English
Toven retweetledi
Toven retweetledi
Perceptron Mk1 is live on OpenRouter, built by @perceptroninc.
Frontier video and embodied reasoning in a vision-language model. Analyzes video at a dynamic frame rate (up to 2 FPS) across a 32k multimodal context, with hybrid reasoning and structured spatial primitives (points, boxes, polygons, clips) as first-class outputs.
English
Toven retweetledi
absolutely unhinged and hilarious behavior in agent royale
@GeminiApp 3 tries to ram grok 4.1.. then proceeds gets out of the car (??)
grok 4.1 takes the opportunity to STEAL gemini's car, and proceeds to ram him to death
.. amazing
English
Toven retweetledi
Trinity-Large-Thinking is now free to use in @OpenRouter through May 23.
This model is a great choice for those working with complex agentic workflows, heavy multi-turn tool use, and long-horizon routing.
The free endpoint is live: openrouter.ai/arcee-ai/trini…
English
Toven retweetledi
I have to say, as someone relatively new to using @OpenRouter, it's a fantastic product. Especially when combined with an open-source harness.
It's very eye-opening to quickly swap between models using the same constructed plan and see the differences in tokens, time, and implementation quality.
English
All the haters who called it a wrapper and will be dead in a few years were so wrong.
It's in this weird space of being a "1 company" where there are no other real rivals.
Hugging Face is another except for Modelscope.
Cody Blakeney@code_star
I have to say, as someone relatively new to using @OpenRouter, it's a fantastic product. Especially when combined with an open-source harness. It's very eye-opening to quickly swap between models using the same constructed plan and see the differences in tokens, time, and implementation quality.
English
@code_star @OpenRouter 🙌 awesome to hear. please don’t hesitate to ping me with feedback though!
English
Toven retweetledi
You can now access the Pareto Code feature from OpenRouter in Hermes Agent, see the docs on use here:
#openrouter-routing--pareto-code-for-auxiliary-tasks" target="_blank" rel="nofollow noopener">hermes-agent.nousresearch.com/docs/user-guid…
OpenRouter@OpenRouter
Introducing Pareto Code: a new, free, experimental coding router Set `min_coding_score` in your request and route to the cheapest code-capable model that clears your bar, ranked by @ArtificialAnlys. See the Pareto frontier shifting in real time👇
English
@thdxr agents love your shit @kitlangton:
"5. Bottom Line
Opencode’s packages/llm is a beautifully architected client-side LLM SDK."
English
we're working on a library to abstract over all the llm providers
there's very few teams that have dealt with the quirks between providers at the scale we have
it's written in effect but will also have a vanilla api
progress is in the opencode repo under packages/llm
English
Toven retweetledi
Exciting launch by OpenRouter that uses Artificial Analysis benchmarks
OpenRouter@OpenRouter
Introducing Pareto Code: a new, free, experimental coding router Set `min_coding_score` in your request and route to the cheapest code-capable model that clears your bar, ranked by @ArtificialAnlys. See the Pareto frontier shifting in real time👇
English