Mirendil's first model Fern: pretrained from scratch with up to 10M-token context
QuanquanGu · x · 2026-10-27
Mirendil has introduced Fern, its first model, pretrained from scratch with a context window of up to 10M tokens. The team notes that's roughly 0.5M lines of code (about Megatron-LM scale), or days of an agent's work, all in context at once. The first version, Fern 0.1, received a short round of post-training.
Related event: Mirendil Debuts Fern, a From-Scratch Model with 10M-Token Context(12 posts)→
More from Models
- OpenAI lists GPT-6.1 Terra: 1.05M context, $1/$5 pricing — but this poster sticks with Opus 5.5 — BLCNYY · 2026-10-27
- GPT-6.1 Terra cuts factual error rate from 22.3% to 4.3%, an ~81% drop — haider1 · 2026-10-27
- GPT-6.1 Terra lands: early users praise speed, value and subagent performance — BenBajarin · 2026-10-27
- Leaked GPT-6.1 Terra benchmarks: cheap but 'boring', still far behind Astra and Claude — Angaisb_ · 2026-10-27
- Artificial Analysis Ships Intelligence Index v4.3.2 With GPT-6.1 Terra Comparison — ArtificialAnlys · 2026-10-27
- GPT-6.1 Terra cuts hallucination rate by 50 points while lifting accuracy 2 points — ArtificialAnlys · 2026-10-27