Measured performance

Provider & model performance

Name: TrustedRouter LLM provider and model speed leaderboard
Creator: TrustedRouter
License: https://www.apache.org/licenses/LICENSE-2.0

Measured time-to-first-token, time-to-first-byte, throughput, and uptime for every LLM provider and model TrustedRouter routes to — continuously sampled, not vendor-claimed.

Last updated 2026-07-03T15:03:44Z

Continuously sampled from TrustedRouter's monitor regions over the 5,000-sample benchmark set — time-to-first-token (TTFT), time-to-first-byte (TTFB), throughput, and success rate measured on real streaming requests, not vendor-claimed. Unsupported route and probe-configuration rows are reported separately and do not count as provider downtime. No prompt or output content is ever stored.

Providers

Ranked by measured p50 time-to-first-token across all of a provider's models in the 5,000-sample benchmark set (27 providers · 709 samples).

#	Provider	Models	p50 TTFT	Throughput	Uptime	Errors	Config excluded	Samples
1	deepinfra	8	1538 ms	—	100.00%	—	—	25
2	friendli	8	1916 ms	—	91.67%	`provider_error` 8%	—	24
3	parasail	13	1953 ms	—	95.00%	`provider_error` 5%	—	20
4	venice	10	2179 ms	—	96.30%	`ReadTimeout` 4%	—	27
5	kimi	2	2787 ms	—	100.00%	—	—	32
6	gemini	6	2834 ms	—	100.00%	—	—	26
7	minimax	6	2904 ms	—	96.97%	`ReadTimeout` 3%	—	33
8	siliconflow	7	2935 ms	—	91.30%	`provider_error` 9%	—	23
9	phala	12	3367 ms	—	84.21%	`provider_error` 16%	—	19
10	zai	11	3562 ms	—	100.00%	—	—	31
11	together	4	3744 ms	—	100.00%	—	—	22
12	baseten	10	4099 ms	—	100.00%	—	—	25
13	tinfoil	6	5446 ms	—	96.77%	`ReadTimeout` 3%	—	31
14	grok	2	5728 ms	—	100.00%	—	—	23
15	openai	12	5830 ms	—	100.00%	—	—	31
16	mistral	8	5929 ms	—	100.00%	—	—	35
17	nebius	11	5992 ms	—	100.00%	—	—	18
18	crusoe	12	6129 ms	—	83.33%	`provider_error` 17%	2 `probe_config_error`	24
19	cerebras	4	6282 ms	—	100.00%	—	—	24
20	gmi	4	7279 ms	—	100.00%	—	—	24
21	fireworks	6	7606 ms	—	100.00%	—	—	28
22	deepseek	2	9205 ms	—	100.00%	—	—	25
23	xiaomi	4	9871 ms	—	42.86%	`provider_error` 57%	—	21
24	lightning	1	9965 ms	—	100.00%	—	—	30
25	novita	23	10034 ms	—	89.66%	`provider_error` 10%	—	29
26	wafer	8	10662 ms	—	52.00%	`provider_error` 48%	—	25
27	anthropic	10	10678 ms	69 tok/s	67.65%	`provider_error` 32%	—	34

Models

Models sampled in the 5,000-sample benchmark set, fastest measured TTFT first. Rows with few samples are marked — more data sharpens the numbers.

#	Model	Provider	p50 TTFT	p95 TTFT	p50 TTFB	Throughput	Uptime	Config excluded	Samples
1	nvidia/nemotron-3-nano-omni-reasoning-30b-a3b limited data	crusoe	715 ms	10150 ms	714 ms	—	66.67%	—	3
2	mistralai/mistral-small-3.2-24b-instruct limited data	mistral	717 ms	3558 ms	716 ms	—	100.00%	—	2
3	google/gemma-4-31b-it limited data	deepinfra	803 ms	17006 ms	803 ms	—	100.00%	—	2
4	NousResearch/Hermes-4-405B limited data	nebius	857 ms	857 ms	856 ms	—	100.00%	—	1
5	google/gemma-4-26b-a4b-it limited data	parasail	860 ms	21233 ms	860 ms	—	100.00%	—	2
6	NousResearch/Hermes-4-70B limited data	nebius	864 ms	864 ms	863 ms	—	100.00%	—	1
7	deepseek/deepseek-v4-flash limited data	crusoe	997 ms	997 ms	997 ms	—	100.00%	—	1
8	qwen/qwen3-vl-235b-a22b-instruct limited data	novita	998 ms	998 ms	998 ms	—	100.00%	—	1
9	qwen/qwen3-vl-8b-instruct limited data	parasail	1006 ms	7963 ms	1005 ms	—	100.00%	—	2
10	google/gemma-3-4b-it limited data	deepinfra	1060 ms	17469 ms	1059 ms	—	100.00%	—	6
11	lgai-exaone/k-exaone-236b-a23b limited data	friendli	1069 ms	9267 ms	1069 ms	—	100.00%	—	3
12	qwen/qwen3-235b-a22b-fp8 limited data	novita	1070 ms	1070 ms	1069 ms	—	50.00%	—	2
13	z-ai/glm-4.7-flash limited data	venice	1124 ms	8940 ms	1123 ms	—	100.00%	—	2
14	z-ai/glm-4.5v limited data	zai	1139 ms	17360 ms	1138 ms	—	100.00%	—	2
15	meta-llama/llama-3.1-8b-instruct limited data	friendli	1146 ms	1146 ms	1145 ms	—	100.00%	—	1
16	deepseek/deepseek-v3.2 limited data	friendli	1268 ms	16387 ms	1268 ms	—	100.00%	—	6
17	inclusionai/ring-2.6-1t limited data	novita	1283 ms	1283 ms	1282 ms	—	100.00%	—	1
18	google/gemma-4-26b-a4b-it limited data	deepinfra	1322 ms	4207 ms	1322 ms	—	100.00%	—	3
19	kwaipilot/kat-coder-pro limited data	novita	1465 ms	1465 ms	1464 ms	—	100.00%	—	1
20	z-ai/glm-5.1 limited data	baseten	1468 ms	3226 ms	1467 ms	—	100.00%	—	3
21	meta-llama/llama-3.1-70b-instruct limited data	deepinfra	1538 ms	11846 ms	1537 ms	—	100.00%	—	2
22	z-ai/glm-4.7 limited data	baseten	1587 ms	14789 ms	1587 ms	—	100.00%	—	4
23	moonshotai/kimi-k2.6 limited data	fireworks	1622 ms	7386 ms	1621 ms	—	100.00%	—	5
24	z-ai/glm-5.1 limited data	venice	1696 ms	8620 ms	1696 ms	—	100.00%	—	2
25	qwen/qwen2.5-vl-72b-instruct limited data	parasail	1714 ms	1714 ms	1713 ms	—	100.00%	—	1
26	nvidia/Llama-3_1-Nemotron-Ultra-253B-v1 limited data	nebius	1783 ms	1783 ms	1782 ms	—	100.00%	—	1
27	tencent/hunyuan-a13b-instruct limited data	siliconflow	1825 ms	4892 ms	1824 ms	—	100.00%	—	3
28	z-ai/glm-5.2 limited data	baseten	1841 ms	4255 ms	1841 ms	—	100.00%	—	2
29	nvidia/nemotron-3-super-120b-a12b limited data	crusoe	1841 ms	1841 ms	1841 ms	—	100.00%	2 `probe_config_error`	1
30	bytedance/ui-tars-1.5-7b limited data	parasail	1857 ms	8426 ms	1857 ms	—	100.00%	—	4
31	openai/gpt-4.1-nano limited data	openai	1890 ms	17591 ms	1889 ms	—	100.00%	—	4
32	google/gemini-2.5-flash limited data	gemini	1911 ms	13973 ms	1910 ms	—	100.00%	—	5
33	minimax/minimax-m2.5 limited data	friendli	1916 ms	7643 ms	1915 ms	—	100.00%	—	3
34	arcee-ai/trinity-large-thinking limited data	parasail	1953 ms	1953 ms	1952 ms	—	100.00%	—	1
35	qwen/qwen3.6-27b limited data	venice	1965 ms	12049 ms	1964 ms	—	100.00%	—	4
36	minimax/minimax-m2.1-highspeed limited data	minimax	1998 ms	3621 ms	1997 ms	—	100.00%	—	7
37	deepseek/deepseek-v3.1-terminus limited data	novita	2013 ms	2013 ms	2012 ms	—	100.00%	—	1
38	qwen/qwen-2.5-7b-instruct limited data	together	2021 ms	18068 ms	2020 ms	—	100.00%	—	6
39	minimax/minimax-m2.7-highspeed limited data	minimax	2037 ms	14266 ms	2036 ms	—	100.00%	—	5
40	qwen/qwen3-235b-a22b-instruct-2507 limited data	novita	2074 ms	2074 ms	2073 ms	—	100.00%	—	1
41	z-ai/glm-5v-turbo limited data	venice	2141 ms	5015 ms	2140 ms	—	100.00%	—	2
42	z-ai/glm-5 limited data	venice	2179 ms	17012 ms	2179 ms	—	100.00%	—	6
43	z-ai/glm-5 limited data	siliconflow	2197 ms	13315 ms	2197 ms	—	100.00%	—	5
44	mistralai/ministral-3b-2512 limited data	mistral	2204 ms	13677 ms	2203 ms	—	100.00%	—	4
45	z-ai/glm-5 limited data	zai	2206 ms	3256 ms	2205 ms	—	100.00%	—	2
46	google/gemini-3.1-flash-lite limited data	gemini	2207 ms	5862 ms	2206 ms	—	100.00%	—	4
47	z-ai/glm-5.2 limited data	zai	2240 ms	2240 ms	2239 ms	—	100.00%	—	1
48	openai/gpt-oss-120b limited data	novita	2291 ms	2291 ms	2290 ms	—	100.00%	—	1
49	moonshotai/kimi-k2.6 limited data	phala	2293 ms	2293 ms	2292 ms	—	100.00%	—	1
50	openai/gpt-oss-120b limited data	phala	2321 ms	5819 ms	2320 ms	—	100.00%	—	2
51	openai/gpt-oss-120b limited data	fireworks	2335 ms	16454 ms	2334 ms	—	100.00%	—	4
52	z-ai/glm-4.6 limited data	zai	2369 ms	3962 ms	2369 ms	—	100.00%	—	2
53	z-ai/glm-5 limited data	baseten	2388 ms	6602 ms	2388 ms	—	100.00%	—	3
54	meta-llama/llama-3.3-70b-instruct limited data	tinfoil	2431 ms	10351 ms	2430 ms	—	100.00%	—	4
55	openai/gpt-4.1-mini limited data	openai	2431 ms	8760 ms	2430 ms	—	100.00%	—	2
56	qwen/qwen3-omni-30b-a3b-instruct limited data	novita	2476 ms	2476 ms	2475 ms	—	100.00%	—	1
57	z-ai/glm-4.7 limited data	venice	2659 ms	2659 ms	2659 ms	—	100.00%	—	1
58	z-ai/glm-5.2 limited data	phala	2681 ms	2681 ms	2680 ms	—	100.00%	—	1
59	moonshotai/kimi-k2.6 limited data	parasail	2779 ms	2779 ms	2778 ms	—	100.00%	—	1
60	moonshotai/kimi-k2.5 limited data	kimi	2787 ms	19478 ms	2786 ms	—	100.00%	—	19
61	mistralai/mistral-large limited data	mistral	2810 ms	13945 ms	2810 ms	—	100.00%	—	2
62	z-ai/glm-4.6v limited data	zai	2830 ms	16667 ms	2829 ms	—	100.00%	—	5
63	google/gemini-2.5-flash-lite limited data	gemini	2834 ms	10319 ms	2833 ms	—	100.00%	—	5
64	deepseek/deepseek-v3-0324 limited data	crusoe	2880 ms	2880 ms	2879 ms	—	100.00%	—	1
65	minimax/minimax-m2 limited data	minimax	2904 ms	11202 ms	2903 ms	—	100.00%	—	6
66	deepseek/deepseek-v4-flash limited data	siliconflow	2935 ms	11913 ms	2934 ms	—	100.00%	—	5
67	minimax/minimax-m2.5-highspeed limited data	minimax	2957 ms	10231 ms	2957 ms	—	100.00%	—	2
68	openai/gpt-oss-120b limited data	cerebras	3063 ms	16691 ms	3063 ms	—	100.00%	—	6
69	mistralai/mistral-small-2603 limited data	mistral	3078 ms	8816 ms	3077 ms	—	100.00%	—	5
70	minimax/minimax-m2.5 limited data	phala	3101 ms	3101 ms	3101 ms	—	100.00%	—	1
71	openai/o4-mini limited data	openai	3154 ms	13986 ms	3154 ms	—	100.00%	—	2
72	moonshotai/kimi-k2.7-code limited data	wafer	3190 ms	14425 ms	3188 ms	—	100.00%	—	3
73	anthropic/claude-opus-4.8 limited data	anthropic	3207 ms	11204 ms	3206 ms	—	100.00%	—	2
74	Qwen/Qwen3-Next-80B-A3B-Thinking limited data	nebius	3346 ms	3346 ms	3346 ms	—	100.00%	—	1
75	moonshotai/kimi-k2.5 limited data	phala	3367 ms	11252 ms	3367 ms	—	80.00%	—	5
76	mistralai/ministral-8b-2512 limited data	mistral	3559 ms	10757 ms	3559 ms	—	100.00%	—	3
77	z-ai/glm-4.7 limited data	zai	3562 ms	15811 ms	3562 ms	—	100.00%	—	4
78	deepseek/deepseek-v3.2 limited data	novita	3646 ms	3646 ms	3646 ms	—	100.00%	—	1
79	z-ai/glm-5.2 limited data	together	3744 ms	12489 ms	3743 ms	—	100.00%	—	6
80	meta-llama/llama-3.3-70b-instruct limited data	crusoe	3763 ms	7789 ms	3763 ms	—	100.00%	—	4
81	qwen/qwen3-coder-next limited data	parasail	3788 ms	3788 ms	3787 ms	—	100.00%	—	1
82	anthropic/claude-sonnet-4.5 limited data	anthropic	3938 ms	8344 ms	3937 ms	—	100.00%	—	2
83	z-ai/glm-5v-turbo limited data	zai	4088 ms	5448 ms	4088 ms	—	100.00%	—	3
84	Sao10K/L3-8B-Stheno-v3.2 limited data	novita	4089 ms	4089 ms	4088 ms	—	100.00%	—	1
85	nvidia/nvidia-nemotron-3-ultra-550b-a55b limited data	baseten	4099 ms	15071 ms	4098 ms	—	100.00%	—	2
86	anthropic/claude-opus-4.6 limited data	anthropic	4100 ms	18298 ms	4100 ms	—	100.00%	—	2
87	z-ai/glm-4.5 limited data	zai	4142 ms	7569 ms	4141 ms	—	100.00%	—	3
88	qwen/qwen3.7-max limited data	wafer	4155 ms	4155 ms	4154 ms	—	100.00%	—	1
89	meta-llama/llama-3.3-70b-instruct limited data	together	4318 ms	13880 ms	4318 ms	—	100.00%	—	5
90	z-ai/glm-4.6 limited data	venice	4444 ms	4444 ms	4443 ms	—	100.00%	—	1
91	google/gemma-4-31b-it limited data	tinfoil	4606 ms	12712 ms	4606 ms	—	85.71%	—	7
92	google/gemini-3-flash-preview limited data	gemini	4633 ms	12792 ms	4632 ms	—	100.00%	—	2
93	anthropic/claude-opus-4.5 limited data	anthropic	4910 ms	9412 ms	4910 ms	—	100.00%	—	3
94	openai/gpt-4o-mini limited data	openai	5053 ms	10008 ms	5052 ms	—	100.00%	—	2
95	deepseek-ai/DeepSeek-V4-Pro limited data	nebius	5150 ms	11021 ms	5149 ms	—	100.00%	—	2
96	deepseek/deepseek-v4-pro limited data	siliconflow	5282 ms	5282 ms	5282 ms	—	100.00%	—	1
97	openai/o3 limited data	openai	5361 ms	15040 ms	5361 ms	—	100.00%	—	4
98	moonshotai/kimi-k2.7-code limited data	baseten	5384 ms	20537 ms	5384 ms	—	100.00%	—	4
99	deepseek/deepseek-v4-pro limited data	tinfoil	5446 ms	11481 ms	5446 ms	—	100.00%	—	5
100	x-ai/grok-4.3 limited data	grok	5641 ms	14745 ms	5639 ms	—	100.00%	—	11
101	x-ai/grok-4.20 limited data	grok	5728 ms	16233 ms	5727 ms	—	100.00%	—	12
102	Qwen/Qwen3-30B-A3B-Instruct-2507 limited data	nebius	5728 ms	5728 ms	5727 ms	—	100.00%	—	1
103	openai/gpt-oss-120b limited data	parasail	5760 ms	6901 ms	5759 ms	—	100.00%	—	3
104	openai/o1 limited data	openai	5830 ms	22327 ms	5830 ms	—	100.00%	—	4
105	deepseek/deepseek-v4-flash limited data	deepseek	5900 ms	30253 ms	5900 ms	—	100.00%	—	12
106	z-ai/glm-5.2 limited data	gmi	5909 ms	16130 ms	5909 ms	—	100.00%	—	8
107	mistralai/ministral-14b-2512 limited data	mistral	5929 ms	14323 ms	5928 ms	—	100.00%	—	10
108	z-ai/glm-4.5-air:free limited data	zai	5937 ms	14366 ms	5936 ms	—	100.00%	—	3
109	openai/gpt-oss-120b limited data	nebius	5992 ms	9396 ms	5992 ms	—	100.00%	—	4
110	z-ai/glm-5.1 limited data	crusoe	6129 ms	16883 ms	6129 ms	—	100.00%	—	2
111	cerebras/zai-glm-4.7 limited data	cerebras	6282 ms	16249 ms	6282 ms	—	100.00%	—	10
112	google/gemma-3-12b-it limited data	deepinfra	6404 ms	14358 ms	6403 ms	—	100.00%	—	5
113	z-ai/glm-5.2 limited data	tinfoil	6470 ms	25624 ms	6470 ms	—	100.00%	—	5
114	z-ai/glm-5v-turbo limited data	siliconflow	6740 ms	14240 ms	6740 ms	—	100.00%	—	2
115	google/gemma-4-31b-it limited data	crusoe	6752 ms	10092 ms	6752 ms	—	100.00%	—	2
116	zai-org/glm-4.6v limited data	novita	6915 ms	6915 ms	6914 ms	—	100.00%	—	1
117	moonshotai/kimi-k2.6 limited data	together	7079 ms	8632 ms	7079 ms	—	100.00%	—	5
118	deepseek/deepseek-v4-pro limited data	gmi	7279 ms	27243 ms	7279 ms	—	100.00%	—	7
119	minimax/minimax-m3 limited data	siliconflow	7311 ms	16744 ms	7310 ms	—	60.00%	—	5
120	qwen/qwen3.5-397b-a17b limited data	phala	7325 ms	7325 ms	7325 ms	—	100.00%	—	1
121	moonshotai/kimi-k2.6 limited data	kimi	7369 ms	24692 ms	7368 ms	—	100.00%	—	13
122	z-ai/glm-4.7 limited data	cerebras	7408 ms	13328 ms	7407 ms	—	100.00%	—	5
123	google/gemma-3-27b-it limited data	nebius	7432 ms	21741 ms	7431 ms	—	100.00%	—	3
124	google/gemini-3.5-flash limited data	gemini	7446 ms	14372 ms	7445 ms	—	100.00%	—	4
125	google/gemini-3.1-flash-lite-preview limited data	gemini	7567 ms	17544 ms	7567 ms	—	100.00%	—	6
126	yutori/n1.5 limited data	crusoe	7590 ms	7590 ms	7589 ms	—	100.00%	—	1
127	z-ai/glm-5.2 limited data	fireworks	7606 ms	19536 ms	7605 ms	—	100.00%	—	6
128	openai/gpt-oss-120b limited data	tinfoil	7700 ms	16717 ms	7699 ms	—	100.00%	—	6
129	qwen/qwen3.5-27b limited data	phala	8213 ms	8213 ms	8212 ms	—	100.00%	—	1
130	z-ai/glm-5.2 limited data	parasail	8376 ms	8376 ms	8376 ms	—	100.00%	—	1
131	mistralai/mistral-nemo limited data	mistral	8707 ms	18132 ms	8706 ms	—	100.00%	—	6
132	minimax/minimax-m2.7 limited data	minimax	8732 ms	15619 ms	8731 ms	—	85.71%	—	7
133	nvidia/nemotron-120b-a12b limited data	baseten	8751 ms	27227 ms	8751 ms	—	100.00%	—	2
134	z-ai/glm-5.1 limited data	friendli	8914 ms	14707 ms	8912 ms	—	100.00%	—	5
135	moonshotai/kimi-k2.6 limited data	tinfoil	9038 ms	28949 ms	9037 ms	—	100.00%	—	4
136	deepseek/deepseek-v4-pro limited data	deepseek	9205 ms	18144 ms	9205 ms	—	100.00%	—	13
137	nvidia/nemotron-3-ultra-550b limited data	crusoe	9222 ms	24256 ms	9221 ms	—	100.00%	—	2
138	z-ai/glm-5.2 limited data	friendli	9268 ms	9268 ms	9267 ms	—	100.00%	—	1
139	xiaomi/mimo-v2.5 limited data	xiaomi	9353 ms	16113 ms	9352 ms	—	100.00%	—	4
140	z-ai/glm-5.1 limited data	fireworks	9442 ms	16683 ms	9433 ms	—	100.00%	—	8
141	cerebras/gpt-oss-120b limited data	cerebras	9503 ms	11306 ms	9502 ms	—	100.00%	—	3
142	deepseek/deepseek-v4-pro limited data	baseten	9542 ms	12429 ms	9541 ms	—	100.00%	—	2
143	zai-org/glm-4.7 limited data	novita	9763 ms	9763 ms	9762 ms	—	100.00%	—	1
144	google/gemma-3-27b-it limited data	deepinfra	9797 ms	10445 ms	9797 ms	—	100.00%	—	3
145	z-ai/glm-5.2 limited data	deepinfra	9828 ms	9828 ms	9828 ms	—	100.00%	—	1
146	xiaomi/mimo-v2.5-pro-ultraspeed limited data	xiaomi	9871 ms	27967 ms	9871 ms	—	100.00%	—	5
147	google/gemma-4-31b-it	lightning	9965 ms	20757 ms	9964 ms	—	100.00%	—	30
148	google/gemma-4-26b-a4b-it limited data	novita	10034 ms	17816 ms	10033 ms	—	100.00%	—	2
149	minimax/minimax-m2.1 limited data	novita	10130 ms	10130 ms	10129 ms	—	100.00%	—	1
150	moonshotai/kimi-k2.5 limited data	fireworks	10338 ms	10338 ms	10338 ms	—	100.00%	—	1
151	minimax/minimax-m3 limited data	wafer	10662 ms	16853 ms	10661 ms	—	100.00%	—	3
152	anthropic/claude-haiku-4.5 limited data	anthropic	10678 ms	13807 ms	10677 ms	—	100.00%	—	5
153	qwen/qwen3.6-35b-a3b limited data	wafer	10770 ms	15516 ms	10769 ms	—	100.00%	—	6
154	openai/gpt-5.4-mini limited data	openai	10790 ms	17874 ms	10789 ms	—	100.00%	—	4
155	qwen/qwen3-235b-a22b-2507 limited data	crusoe	10810 ms	17013 ms	10810 ms	—	100.00%	—	3
156	qwen/qwen3-235b-a22b-2507 limited data	friendli	11105 ms	11982 ms	11105 ms	—	100.00%	—	3
157	nvidia/nemotron-3-super-120b-a12b limited data	nebius	11109 ms	11109 ms	11109 ms	—	100.00%	—	1
158	minimax/minimax-m3 limited data	minimax	11156 ms	14917 ms	11155 ms	—	100.00%	—	6
159	Qwen/Qwen3-235B-A22B-Instruct-2507 limited data	nebius	11391 ms	18231 ms	11390 ms	—	100.00%	—	2
160	minimax/minimax-m2 limited data	novita	11566 ms	11566 ms	11565 ms	—	100.00%	—	1
161	baidu/ernie-4.5-vl-424b-a47b limited data	novita	11748 ms	21939 ms	11747 ms	—	100.00%	—	3
162	z-ai/glm-5-turbo limited data	venice	12048 ms	26285 ms	12048 ms	—	100.00%	—	2
163	z-ai/glm-4.5-air limited data	zai	12453 ms	22696 ms	12453 ms	—	100.00%	—	5
164	qwen/qwen3-30b-a3b-instruct-2507 limited data	phala	12574 ms	15671 ms	12573 ms	—	100.00%	—	2
165	qwen/qwen3.5-9b limited data	venice	12627 ms	12627 ms	12626 ms	—	100.00%	—	1
166	openai/gpt-5.5 limited data	openai	12718 ms	13556 ms	12718 ms	—	100.00%	—	3
167	thedrummer/skyfall-36b-v2 limited data	parasail	12806 ms	12806 ms	12805 ms	—	100.00%	—	1
168	openai/gpt-4o limited data	openai	13035 ms	13035 ms	13034 ms	—	100.00%	—	1
169	z-ai/glm-5.2 limited data	venice	13061 ms	16141 ms	13060 ms	—	83.33%	—	6
170	anthropic/claude-sonnet-4.6 limited data	anthropic	13080 ms	17113 ms	13079 ms	—	100.00%	—	2
171	meta-llama/llama-4-scout-17b-16e-instruct limited data	novita	13160 ms	13703 ms	13159 ms	—	100.00%	—	3
172	deepseek/deepseek-v4-pro limited data	fireworks	13273 ms	15099 ms	13272 ms	—	100.00%	—	4
173	mistralai/mistral-medium-3-5 limited data	mistral	13450 ms	17254 ms	13450 ms	—	100.00%	—	3
174	z-ai/glm-5.2 limited data	siliconflow	13577 ms	24924 ms	13577 ms	—	100.00%	—	2
175	openai/o3-mini limited data	openai	13761 ms	14415 ms	13761 ms	—	100.00%	—	3
176	moonshotai/kimi-k2.5 limited data	baseten	13881 ms	14537 ms	13881 ms	—	100.00%	—	2
177	qwen/qwen-mt-plus limited data	novita	13900 ms	13900 ms	13900 ms	—	100.00%	—	1
178	qwen/qwen3.5-27b limited data	deepinfra	13981 ms	15164 ms	13980 ms	—	100.00%	—	3
179	z-ai/glm-5.1 limited data	gmi	14167 ms	16611 ms	14167 ms	—	100.00%	—	7
180	anthropic/claude-opus-4.7 limited data	anthropic	14599 ms	18186 ms	14598 ms	69 tok/s	100.00%	—	7
181	openai/gpt-oss-120b limited data	crusoe	14896 ms	14896 ms	14895 ms	—	100.00%	—	1
182	z-ai/glm-5.1 limited data	zai	15113 ms	15113 ms	15113 ms	—	100.00%	—	1
183	openai/gpt-oss-20b limited data	parasail	15700 ms	15700 ms	15700 ms	—	100.00%	—	1
184	openai/gpt-oss-20b limited data	phala	15711 ms	15711 ms	15711 ms	—	100.00%	—	1
185	zai-org/GLM-5.1 limited data	nebius	16398 ms	16398 ms	16397 ms	—	100.00%	—	1
186	mistralai/mistral-small-3.2-24b-instruct limited data	parasail	16761 ms	16761 ms	16760 ms	—	100.00%	—	1
187	qwen/qwen3-vl-235b-a22b-thinking limited data	novita	16772 ms	16772 ms	16772 ms	—	100.00%	—	1
188	z-ai/glm-5 limited data	gmi	17286 ms	35532 ms	17285 ms	—	100.00%	—	2
189	openai/gpt-oss-120b limited data	baseten	18011 ms	18011 ms	18011 ms	—	100.00%	—	1
190	qwen/qwen-2.5-7b-instruct limited data	phala	18016 ms	18016 ms	18015 ms	—	100.00%	—	1
191	openai/gpt-4.1 limited data	openai	19886 ms	19886 ms	19886 ms	—	100.00%	—	1
192	qwen/qwen3.5-122b-a10b limited data	novita	20610 ms	20610 ms	20609 ms	—	100.00%	—	1
193	deepseek/deepseek-v3.2 limited data	phala	22315 ms	22315 ms	22315 ms	—	100.00%	—	1
194	deepseek/deepseek-r1-turbo limited data	novita	23584 ms	23584 ms	23583 ms	—	100.00%	—	1
195	z-ai/glm-4.7-flash limited data	phala	—	—	—	—	0.00%	—	2
196	anthropic/claude-opus-4 limited data	anthropic	—	—	—	—	0.00%	—	4
197	meta-llama/llama-3.3-70b-instruct limited data	friendli	—	—	—	—	0.00%	—	2
198	qwen/qwen3.5-397b-a17b limited data	wafer	—	—	—	—	0.00%	—	2
199	openai/text-embedding-3-large limited data	openai	—	—	—	—	100.00%	—	1
200	xiaomi/mimo-v2-flash limited data	xiaomi	—	—	—	—	0.00%	—	6
201	anthropic/claude-opus-4.1 limited data	anthropic	—	—	—	—	0.00%	—	4
202	xiaomi/mimo-v2-pro limited data	xiaomi	—	—	—	—	0.00%	—	6
203	deepseek/deepseek-v4-pro limited data	wafer	—	—	—	—	0.00%	—	5
204	zai-org/glm-4.5 limited data	novita	—	—	—	—	0.00%	—	1
205	anthropic/claude-sonnet-4 limited data	anthropic	—	—	—	—	0.00%	—	3
206	deepseek/deepseek-v4-flash limited data	wafer	—	—	—	—	0.00%	—	2
207	moonshotai/kimi-k2.6 limited data	wafer	—	—	—	—	0.00%	—	3
208	moonshotai/kimi-k2.6 limited data	crusoe	—	—	—	—	0.00%	—	3
209	google/gemma-4-31b-it limited data	parasail	—	—	—	—	0.00%	—	1
210	baidu/ernie-4.5-21B-a3b limited data	novita	—	—	—	—	0.00%	—	1