Mistral and NVIDIA's Nemotron Coalition will train and release a new open-source frontier model.
“The benefit for everyone involved, really, is that we will have a new open source frontier model that everyone can build off on.”
open-source-models
9 tracked signals on open-source-models.
Microsoft Foundry introduces Foundry Managed Compute, a managed PaaS for deploying open-source and custom models.
“We have this construct called foundry managed compute which is think of it as a managed platform as a service where you just pick the GPUs.”
Cast AI's Kimchi agent routes tasks to cheap open-source models to eliminate developer token rationing
“Our job is not to prohibit a developer from using an agent for coding. Our job is to make it so they can use it as much as they want, whenever they want, without any restrictions.”
Qwen 3.8 27B defaults to extreme overthinking, consuming 22K tokens for simple tasks
“This is a hilarious default. It's absolutely not a good way to run the model, especially on consumer hardware.”
Decagon runs 90% open-source models because they outperform frontier models on specific enterprise tasks
“An AI agent should just be the front door of your business.”
Open source models still trail closed models on general agentic tasks but can outperform them once fine-tuned for a specific domain.
“because they're open and because you own the weights, you can actually fine-tune or RL them on your specific domain, which can actually make them perform better than those same closed models”
Post-training open-source reasoning models with reinforcement learning keeps agent costs flat while quality improves in Foundry.
“deploying a model into production is just the beginning”
Fine-tuned small open-source reasoning models on Foundry can match frontier models for your domain at a fraction of the cost.
“This is almost like test-driven development for the age of agents.”
Fireworks AI brings day-zero open-source model inference to Microsoft Foundry for production AI workflows.
“We serve around 13 trillion tokens per day and 180,000 requests per second.”