Guess what.
Qwen3.8-Max should already work with Raven π¦ββ¬
Raven has native DashScope support plus custom OpenAI-compatible endpoints, including streaming, tool calls, and reasoning_content.
github.com/EverMind-AI/raβ¦
π’Meet Qwen3.8-Max β our most capable model to date.
Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!π
Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters:
- Autonomous coding: 10+ days of
The LLM can propose changes. But deterministic evals have to decide what deserves credit.
That diagnose-and-credit loop is what turns self-improvement from a demo into a system you can trust.
Our new paper.
Self-evolving agents have a measurement problem.
The hard part isnβt generating changes to an agent harness. Itβs knowing which changes genuinely helped rather than got lucky, overfit, or never activated at all.
LLMs diagnose failures and propose changes across