Ollama
Get up and running with Kimi-K2.6, GLM-5.1, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
Ollama scores 7.8/10 (Strong) in Inference & Serving.
Language Go License MIT Stars 175k★ Type lightweight Self-host Yes Repo ollama/ollama Homepage https://ollama.com
Quality
Strong 7.8
Adoption 81
Activity 80
Maturity 96
Community 81
Capability 2
Show the math
Overall 7.8/10
value = 7.8/10
score = value
Adoption 8.3/10
value = 81.1/100
score = 1 + 9 * value / 100
Activity 8.2/10
value = 80.1/100
score = 1 + 9 * value / 100
Maturity 9.7/10
value = 96.2/100
score = 1 + 9 * value / 100
Community 8.3/10
value = 81.3/100
score = 1 + 9 * value / 100
Capability 1.2/10
value = 2/100
score = 1 + 9 * value / 100
Key metrics
175.4k★ GitHub stars
246/90d Recent commits
16 devs/90d Recent contributors
<2 tok/s peak Decode throughput under concurrency (Qwen3-4B, H100 80GB) MDPI peer-reviewed benchmark (Appl. Sci. 2026, Qwen3-4B, single H100 80GB): Ollama stayed below ~2 tok/s decode rate across all concurrency levels while vLLM reached up to 2891 tok/s; Ollama hit a bottleneck near 10 concurrent users. Higher=better, so Ollama ranks lowest among serving backends. Source: https://www.mdpi.com/2076-3417/16/11/5435
3y Project age
active Last commit
Gotchas
No gotchas documented yet. Contribute one if you know a constraint we missed.
Labels
- Self-hosted
Self-hosted
You deploy and operate it yourself; there is no hosted option here.
local cli openai-compatible gguf desktop