Provider latency

Median provider latency across OpenRouter endpoints, in seconds (3d).

180 models · 180 results · data as of 2026-10-29 · openrouter.ai ↗

All results

180 rows

#	Model	Effort	Provider latency
1	Llama 3.2 1b	none	0.2s
2	Qwen 2.5 7B		0.2s
3	Codestral 2508		0.3s
4	Llama 3.2 3b	none	0.3s
5	Mistral Small 3	none	0.4s
6	Mistral Medium 3	none	0.4s
7	Mistral Small 3.1	none	0.4s
8	Llama 3.1 70b	none	0.4s
9	Llama 3.1 8b	medium	0.4s
10	Inkling	max	0.5s
11	gpt-oss-120b	high	0.5s
12	Mistral	none	0.5s
13	Mistral Small 4	high	0.5s
14	Qwen 3 235B-A22B	xhigh	0.5s
15	Qwen 3.5 Flash	max	0.5s
16	Gemini 2.5 Flash Lite	none	0.5s
17	gpt-oss-20b	high	0.5s
18	Gemma 4 26B	xhigh	0.5s
19	Mistral Large 2407		0.5s
20	Gemma 2 27B		0.6s
21	Llama 4 Maverick	medium	0.6s
22	Haiku 4.5	xhigh	0.6s
23	GPT-3.5	high	0.6s
24	DeepSeek V4.1 Flash	max	0.6s
25	Grok 4.20	xhigh	0.6s
26	Qwen 3.6 Flash	max	0.6s
27	Qwen 2.5 72B	none	0.6s
28	Mistral Large	none	0.6s
29	Mistral Medium 3.1	none	0.6s
30	Llama 3.3 70b	none	0.6s