StackMap
Subscribe

colibri vs Ollama

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 — Run Llama, Mistral and other open models locally with a single command and a clean API.

The curated verdict

Both run open models locally. Ollama is the multi-model daily driver for models that fit; colibrì is a single-model specialist that makes a 744B MoE fit where nothing else will.

colibriOllama
Stars17k176k
Forks1.5k17k
LanguageCGo
LicenseApache-2.0MIT
Last activitytoday2 days ago
Topicslocallocal
Curated connections320

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.

Ollama — the curator's take

Run Llama, Mistral and other open models locally with a single command and a clean API.