colibri vs Strata
Pure-C, zero-dep MoE runtime that runs GLM-5.2 (744B) on a 25GB-RAM consumer box by streaming experts from disk — VRAM/RAM/NVMe as one tiered hierarchy, never touching precision. — versus — Hybrid GPU/CPU MoE engine that runs Qwen3.8-Flash-Next and DeepSeek V4.1 on consumer cards: hot experts cached in VRAM, rest computed in RAM, MTP drafting, local OpenAI/Anthropic API.
Both run MoE models far bigger than VRAM on consumer boxes by tiering experts across VRAM, RAM and disk; Colibri is a zero-dep C runtime that streams from NVMe, Strata keeps experts in RAM and computes them on the CPU alongside the GPU.
| colibri | Strata | |
|---|---|---|
| Stars | 48k | 41k |
| Forks | 5.6k | 4.4k |
| Language | C | C++ |
| License | Apache-2.0 | MIT |
| Last activity | yesterday | yesterday |
| Topics | local | local, coding |
| Curated connections | 7 | 5 |
colibri — the curator's take
Use it when you want a frontier-scale open MoE (GLM-5.2, 744B) on hardware that can't hold it, and you care about fidelity: forward pass is token-exact against the transformers oracle, and placement only ever changes speed, never semantics. The learning cache pins the experts YOUR workload actually routes, so it genuinely gets faster with use. NOT a general runner — it's a one-model engine; for everyday multi-model local use pick ollama. And expect 0.05–6.8 tok/s depending on tier residency: this is patience-ware for correctness fanatics, not latency-sensitive serving.
Strata — the curator's take
Reach for it when you own a 12-24 GB GPU (NVIDIA, AMD, Apple Silicon) and want a frontier-class MoE behind Claude Code, Codex or OpenCode with no cloud — one installer, `strata connect` wires the tools. It is a purpose-built engine for a handful of model families, not a general runner: for arbitrary GGUFs or multi-user serving use Ollama/vLLM, and on low RAM or small disks the DeepSeek packs (200+ GB) will not fit.