Prime Intellect Releases Prime Agent: An Open-Source RLM Harness Where Sub-Agents Are Function Calls Inside Persistent IPython Kernel
Prime Intellect has open-sourced Prime Agent, a self-improving coding harness designed around two abstractions, the Recursive Language Model (RLM) and Continual Harness. Fixed tool schemas and context compaction force a model to work around its own scaffolding. Prime Agent replaces both with a persistent Python REPL and a rewritable harness. With Opus 5, it reports 95.5% on ARC-AGI-3, above the reported human expert baseline of 95.4%. It is MIT-licensed. Is it deployable Yes, today. Prime
Get smart on it
Prime Agent is an open-source tool that helps AI language models complete complex coding tasks by giving them access to a persistent Python environment where they can write and execute code. Rather than using fixed tool schemas, the system treats sub-agents as function calls within this environment, allowing the model to create and manage its own skills and memory. The tool is designed for engineering organizations and AI research labs working on multi-hour tasks, and can run on various language models through subscriptions, APIs, or self-hosted options. In benchmarks, Prime Agent achieved above-human performance on a reasoning task, reaching 95.5% accuracy compared to a human expert baseline of 95.4%.
Related news
Pokee AI Releases Pokee-Isaac 28B: A 10M-Token Context Agentic Model Built to Run Inside the Customer Boundary
Long-horizon agents accumulate context faster than they resolve tasks. Every tool output, observation, and intermediate reasoning step stays in the window, and the two capabilities that matter — holding that context and staying coherent across it — have so far been available almost exclusively from cloud endpoints. That excludes regulated industries, public-sector institutions, and on-device applications, where the data is not permitted to leave the boundary at all. Pokee AI released Pokee-Isaa
Mistral AI Releases Shieldstral 1.0 3B: An Open-Weights Policy-Adaptive Multimodal Safety Classifier Matching Models 7× Its Size
Mistral AI has released Shieldstral 1.0 3B, an open-weights, policy-adaptive multimodal safety classifier that treats content moderation as a single yes/no question rather than a fixed taxonomy of harm categories. Most guardrail models bake their category list into the weights, so re-targeting one to a new deployment context means retraining — and the same content can be acceptable on a cybersecurity research tool while being harmful on a mental-health platform. Shieldstral inverts that: operat
