@maximelabonne: Run wild and destroy our serving infra
Summary
Liquid AI's D1 decision model is now live on OpenRouter, offering typed probabilistic answers for state-based decision queries at $0.04/M input with zero data retention. The community is hyped to stress-test OpenRouter's serving infrastructure with the new model.
View Cached Full Text
Cached at: 10/02/26, 10:39 AM
Run wild and destroy our serving infra 🔥
OpenRouter (@OpenRouter): D1 from @liquidai is live on OpenRouter.
It’s a competitive decision model: send your app’s state and yes/no, choice, or score questions, and get back typed answers with a probability for every option. Zero data retention.
$0.04/M input, $0 output, 65K context
Similar Articles
@maximelabonne: LFM2.5-2.6B and LFM2.5-Embedding-350M are now available for free on @OpenRouter
Two AI models, LFM2.5-2.6B and LFM2.5-Embedding-350M, are now available for free on the OpenRouter platform.
@alexatallah: If you're a researcher looking to: → conduct rigorous studies on how multiple models can outperform the frontier → leve…
OpenRouter launches Fusion API, a compound model that achieves high intelligence at half the price, leveraging the largest LLM marketplace.
@pirroh: https://x.com/pirroh/status/2104976589180252218
Replit shares its 'free the models' harness design, where the core agent loop — not an external router — chooses subagent tiers, effort, and delegation as tasks unfold. Using GPT-6 Astra, the approach reportedly beats router-based baselines and a sidekick architecture on DeepSWE and Terminal-Bench while keeping quality at minimum cost.
@maximelabonne: Nature healed: LFM2.5-2.6B is free on OpenRouter and has good throughput! Try it in your favorite agentic harness today.
The LFM2.5-2.6B AI model is free to use on OpenRouter with good throughput, and is promoted for agentic harness applications.
Anyone tried new free (for a week) 1Tmodel on openrouter? how is ring-2.6-1T fit in real work?
The article discusses the new Ring-2.6-1T model on OpenRouter, highlighting its adaptive reasoning capabilities and suitability for coding agents and complex workflows.