Axion AI
AI developments only — models, agents, harnesses, research and the open ecosystem.
Latest
- OpenAI's GPT-6 Astra hits 87.4% on MultiLangBench, but the COBOL migration eval is already near saturation
- Hermes Agent vs Claude Code vs OpenCode: which coding agent should run your terminal in late 2026?
- TorchNeuron explained: Hugging Face says 89% of top text-generation architectures now run natively on AWS Trainium
Models & Releases
-
OpenAI's GPT-6 Astra hits 87.4% on MultiLangBench, but the COBOL migration eval is already near saturation
OpenAI says its newest models jump from 26.8% to 87.4% on a 20-language benchmark and from 5.4% to 86.6% on COBOL migration. The numbers are real; the sample sizes are not published.
Agents & Harnesses
-
Hermes Agent vs Claude Code vs OpenCode: which coding agent should run your terminal in late 2026?
Three agents, three philosophies: a self-improving generalist with a messaging gateway, a vendor-tied CLI shipping daily, and an MIT-licensed coding TUI. An Axion Scorecard on what we can actually verify.
Open Ecosystem
-
TorchNeuron explained: Hugging Face says 89% of top text-generation architectures now run natively on AWS Trainium
A new PyTorch backend replaces the XLA-based Optimum Neuron path. Here is what changes for practitioners, what the numbers actually say, and the static-shape catch.