LMSpeed Model Speed Leaderboard
Multi-dimensional rankings based on model speed tests, provider health checks, and standard model benchmarks. Compare providers, endpoints, models, and reliability at a glance.
Compares AI model responsiveness across Artificial Analysis output speed, time to first token, and time to first answer token in one table.
Artificial Analysis standard benchmarks
Fairness Notice
These model leaderboards use imported Artificial Analysis benchmark records and are deduplicated to one canonical LMSpeed model per rank.
| Rank | Model | Output Speed | Time to First Token | Time to First Answer Token | Updated |
|---|---|---|---|---|---|
1 | Mercury 2mercury-2 | 1011.9 tok/s#1 | 4.14 s#59 | 4.14 s#24 | Sep 12, 2026, 04:45 AM |
2 | Trinity Large Thinkingtrinity-large-thinking | 365.5 tok/s#2 | 1.05 s#32 | 6.53 s#25 | Sep 12, 2026, 04:45 AM |
3 | Gemini 3.5 Flash-Litegemini-3-5-flash-lite | 345.0 tok/s#3 | 7.71 s#64 | 7.71 s#29 | Sep 12, 2026, 04:45 AM |
4 | Gemini 3.8 Flashgemini-3-8-flash | 339.0 tok/s#4 | 12.31 s#69 | 12.31 s#37 | Sep 12, 2026, 04:45 AM |
5 | Ling-3.0-flashling-3-0-flash | 333.9 tok/s#5 | 1.72 s#48 | 7.70 s#28 | Sep 12, 2026, 04:45 AM |
6 | Gemini 3.7 Flashgemini-3-7-flash | 330.5 tok/s#6 | 8.51 s#66 | 8.51 s#30 | Sep 12, 2026, 04:45 AM |
7 | Muse Spark 1.3muse-spark-1-3 | 324.0 tok/s#7 | 28.08 s#74 | 34.25 s#62 | Sep 12, 2026, 04:45 AM |
8 | Nemotron 3.5 Lightningnemotron-3-5-lightning | 285.4 tok/s#8 | 0.56 s#11 | 7.56 s#27 | Sep 12, 2026, 04:45 AM |
9 | DeepSeek V4.1 Flashdeepseek-v4-1-flash | 260.5 tok/s#9 | 0.92 s#25 | 8.60 s#32 | Sep 12, 2026, 04:45 AM |
10 | Muse Spark 1.2muse-spark-1-2 | 250.1 tok/s#10 | 10.81 s#67 | 18.81 s#49 | Sep 12, 2026, 04:45 AM |
11 | gpt-oss-20bgpt-oss-20b | 249.2 tok/s#11 | 0.53 s#10 | 8.55 s#31 | Sep 12, 2026, 04:45 AM |
12 | Gemini 3.5 Flashgemini-3-5-flash | 246.7 tok/s#12 | 0.81 s#20 | 0.81 s#12 | Sep 12, 2026, 04:45 AM |
13 | Qwen3.5 Omni Flashqwen3-5-omni-flash | 241.8 tok/s#13 | 0.94 s#26 | 0.94 s#15 | Sep 12, 2026, 04:45 AM |
14 | DeepSeek V4 Flashdeepseek-v4-flash | 236.6 tok/s#14 | 0.80 s#19 | 9.26 s#33 | Sep 12, 2026, 04:45 AM |
15 | Gemini 3.6 Flashgemini-3-6-flash | 215.8 tok/s#15 | 13.35 s#70 | 13.35 s#38 | Sep 12, 2026, 04:45 AM |
16 | gpt-oss-120bgpt-oss-120b | 196.7 tok/s#16 | 0.50 s#9 | 10.67 s#34 | Sep 12, 2026, 04:45 AM |
17 | Ling 3.0 Tinyling-3-0-tiny | 196.3 tok/s#17 | 2.08 s#55 | 12.27 s#36 | Sep 12, 2026, 04:45 AM |
18 | Qwen3 Next 80B A3B Instructqwen3-next-80b-a3b-instruct | 194.6 tok/s#18 | 0.99 s#28 | 0.99 s#17 | Sep 12, 2026, 04:45 AM |
19 | Nemotron 3 Ultranemotron-3-ultra-550b-a55b | 180.1 tok/s#19 | 1.26 s#39 | 13.89 s#39 | Sep 12, 2026, 04:45 AM |
20 | Qwen3.5-35B-A3Bqwen3-5-35b-a3b | 160.4 tok/s#20 | 1.38 s#42 | 1.38 s#22 | Sep 12, 2026, 04:45 AM |
21 | Solar Pro 3solar-pro-3 | 157.2 tok/s#21 | 1.32 s#41 | 14.05 s#41 | Sep 12, 2026, 04:45 AM |
22 | Inkling Smallinkling-small | 156.9 tok/s#22 | 1.90 s#52 | 14.64 s#42 | Sep 12, 2026, 04:45 AM |
23 | Devstral 2devstral-2 | 149.5 tok/s#23 | 0.72 s#17 | 0.72 s#11 | Sep 12, 2026, 04:45 AM |
24 | Qwen3.5-122B-A10Bqwen3-5-122b-a10b | 145.9 tok/s#24 | 1.03 s#29 | 1.03 s#18 | Sep 12, 2026, 04:45 AM |
25 | Ling 3.0 Flash VLling-3-0-flash-vl | 143.4 tok/s#25 | 1.69 s#47 | 15.64 s#44 | Sep 12, 2026, 04:45 AM |
26 | Mistral Medium 3.5mistral-medium-3-5 | 142.2 tok/s#26 | 0.74 s#18 | 14.81 s#43 | Sep 12, 2026, 04:45 AM |
27 | Qwen3.6 35B A3Bqwen3-6-35b-a3b | 141.0 tok/s#27 | 1.05 s#31 | 39.31 s#66 | Sep 12, 2026, 04:45 AM |
28 | GPT-5.3 Codexgpt-5-3-codex | 138.6 tok/s#28 | 41.56 s#78 | 41.56 s#69 | Sep 12, 2026, 04:45 AM |
29 | Granite 4.1 8Bgranite-4-1-8b | 133.7 tok/s#29 | 0.43 s#7 | 0.43 s#5 | Sep 12, 2026, 04:45 AM |
30 | Step 3.7 Flashstep-3-7-flash | 132.8 tok/s#30 | 1.58 s#45 | 16.64 s#45 | Sep 12, 2026, 04:45 AM |
31 | Ring 2 6 1tring-2-6-1t | 127.1 tok/s#31 | 2.02 s#53 | 17.75 s#47 | Sep 12, 2026, 04:45 AM |
32 | Nemotron 3 Supernemotron-3-super-120b-a12b | 127.1 tok/s#32 | 4.30 s#60 | 20.04 s#51 | Sep 12, 2026, 04:45 AM |
33 | Nex n2 Pronex-n2-pro | 127.1 tok/s#32 | 1.23 s#37 | 16.96 s#46 | Sep 12, 2026, 04:45 AM |
34 | Gemini 3.1 Pro Previewgemini-3-1-pro-preview | 125.2 tok/s#34 | 24.48 s#73 | 24.48 s#56 | Sep 12, 2026, 04:45 AM |
35 | GPT-5.6 Lunagpt-5-6-luna | 121.3 tok/s#35 | 96.08 s#80 | 96.08 s#79 | Sep 12, 2026, 04:45 AM |
36 | Qwen3 Coder Nextqwen3-coder-next | 117.2 tok/s#36 | 1.03 s#30 | 1.03 s#19 | Sep 12, 2026, 04:45 AM |
37 | O3o3 | 114.5 tok/s#37 | 6.58 s#62 | 6.58 s#26 | Sep 12, 2026, 04:45 AM |
38 | Grok 4.3grok-4-3 | 109.2 tok/s#38 | 0.86 s#23 | 0.86 s#14 | Sep 12, 2026, 04:45 AM |
39 | GLM 5.3 Flashglm-5-3-flash | 108.2 tok/s#39 | 2.04 s#54 | 20.53 s#52 | Sep 12, 2026, 04:45 AM |
40 | MiniMax M3minimax-m3 | 106.4 tok/s#40 | 0.83 s#22 | 19.63 s#50 | Sep 12, 2026, 04:45 AM |
41 | Llama 4 Scoutllama-4-scout | 99.6 tok/s#41 | 0.69 s#16 | 0.69 s#10 | Sep 12, 2026, 04:45 AM |
42 | Llama 4 Maverickllama-4-maverick | 96.3 tok/s#42 | 0.57 s#13 | 0.57 s#7 | Sep 12, 2026, 04:45 AM |
43 | North Mini Codenorth-mini-code | 95.6 tok/s#43 | 0.21 s#1 | 21.14 s#53 | Sep 12, 2026, 04:45 AM |
44 | Granite 4.2 8Bgranite-4-2-8b | 95.3 tok/s#44 | 0.25 s#2 | 21.25 s#54 | Sep 12, 2026, 04:45 AM |
45 | GPT-5.6 Terragpt-5-6-terra | 94.9 tok/s#45 | 0.83 s#21 | 0.83 s#13 | Sep 12, 2026, 04:45 AM |
46 | Claude Haiku 4.5claude-haiku-4-5 | 92.1 tok/s#46 | 0.63 s#14 | 0.63 s#8 | Sep 12, 2026, 04:45 AM |
47 | Ministral 3ministral-3 | 91.6 tok/s#47 | 0.56 s#12 | 0.56 s#6 | Sep 12, 2026, 04:45 AM |
48 | Qwen3.5 Omni Plusqwen3-5-omni-plus | 90.4 tok/s#48 | 1.21 s#36 | 1.21 s#21 | Sep 12, 2026, 04:45 AM |
49 | Qwen3.5-9Bqwen3-5-9b | 86.3 tok/s#49 | 0.43 s#6 | 0.43 s#4 | Sep 12, 2026, 04:45 AM |
50 | Hy3hy3 | 86.0 tok/s#50 | 1.72 s#49 | 24.96 s#57 | Sep 12, 2026, 04:45 AM |
51 | Reka Flash 3reka-flash-3 | 85.9 tok/s#51 | 0.88 s#24 | 24.17 s#55 | Sep 12, 2026, 04:45 AM |
52 | Inklinginkling | 83.3 tok/s#52 | 2.63 s#56 | 26.63 s#58 | Sep 12, 2026, 04:45 AM |
53 | Claude Sonnet 5claude-sonnet-5 | 83.2 tok/s#53 | 114.22 s#81 | 114.22 s#81 | Sep 12, 2026, 04:45 AM |
54 | Qwen3.5 397B A17Bqwen3-5-397b-a17b | 82.6 tok/s#54 | 1.61 s#46 | 1.61 s#23 | Sep 12, 2026, 04:45 AM |
55 | Mistral Large 3mistral-large-3 | 81.3 tok/s#55 | 0.68 s#15 | 0.68 s#9 | Sep 12, 2026, 04:45 AM |
56 | GLM-5.2glm-5-2 | 72.3 tok/s#56 | 7.60 s#63 | 35.26 s#63 | Sep 12, 2026, 04:45 AM |
57 | Grok 4.6grok-4-6 | 71.4 tok/s#57 | 38.01 s#77 | 38.01 s#65 | Sep 12, 2026, 04:45 AM |
58 | Qwen3.7 Plusqwen3-7-plus | 70.1 tok/s#58 | 1.46 s#44 | 29.97 s#59 | Sep 12, 2026, 04:45 AM |
59 | Claude Fable 5claude-fable-5 | 70.0 tok/s#59 | 60.71 s#79 | 60.71 s#77 | Sep 12, 2026, 04:45 AM |
60 | Command Acommand-a | 68.6 tok/s#60 | 0.32 s#3 | 0.32 s#1 | Sep 12, 2026, 04:45 AM |
61 | DeepSeek V4 Prodeepseek-v4-pro | 68.4 tok/s#61 | 1.07 s#33 | 30.18 s#60 | Sep 12, 2026, 04:45 AM |
62 | GPT-5.6 Solgpt-5-6-sol | 66.1 tok/s#62 | 17.83 s#72 | 17.83 s#48 | Sep 12, 2026, 04:45 AM |
63 | Claude Opus 5claude-opus-5 | 61.4 tok/s#63 | 36.03 s#76 | 36.03 s#64 | Sep 12, 2026, 04:45 AM |
64 | Grok 4.5grok-4-5 | 58.4 tok/s#64 | 13.91 s#71 | 13.91 s#40 | Sep 12, 2026, 04:45 AM |
65 | Claude Fable 5.1claude-fable-5-1 | 57.8 tok/s#65 | 11.31 s#68 | 11.31 s#35 | Sep 12, 2026, 04:45 AM |
66 | Qwen3.6 27Bqwen3-6-27b | 56.9 tok/s#66 | 1.45 s#43 | 101.17 s#80 | Sep 12, 2026, 04:45 AM |
67 | GPT-6 Astragpt-6-astra | 56.3 tok/s#67 | 32.54 s#75 | 32.54 s#61 | Sep 12, 2026, 04:45 AM |
68 | Claude Sonnet 4.6claude-sonnet-4-6 | 55.3 tok/s#68 | 1.14 s#35 | 1.14 s#20 | Sep 12, 2026, 04:45 AM |
69 | GLM 5.3glm-5-3 | 54.8 tok/s#69 | 3.57 s#58 | 40.10 s#67 | Sep 12, 2026, 04:45 AM |
70 | MiMo-V2.5mimo-v2-5 | 51.7 tok/s#70 | 6.13 s#61 | 44.82 s#70 | Sep 12, 2026, 04:45 AM |
71 | Solar Pro 4solar-pro4 | 50.1 tok/s#71 | 1.09 s#34 | 41.00 s#68 | Sep 12, 2026, 04:45 AM |
72 | Phi 4 Mini Instructphi-4-mini-instruct | 47.5 tok/s#72 | 0.36 s#4 | 0.36 s#2 | Sep 12, 2026, 04:45 AM |
73 | MiMo-V2.5-Promimo-v2-5-pro | 44.4 tok/s#73 | 7.72 s#65 | 52.72 s#74 | Sep 12, 2026, 04:45 AM |
74 | Phi 4phi-4 | 43.7 tok/s#74 | 0.95 s#27 | 0.95 s#16 | Sep 12, 2026, 04:45 AM |
75 | Qwen3.8 27Bqwen3-8-27b | 43.1 tok/s#75 | 1.31 s#40 | 47.66 s#71 | Sep 12, 2026, 04:45 AM |
76 | Kimi K2.7 Codekimi-k2-7-code | 42.5 tok/s#76 | 1.25 s#38 | 53.67 s#75 | Sep 12, 2026, 04:45 AM |
77 | Qwen3.8 Maxqwen3-8-max | 41.4 tok/s#77 | 1.80 s#50 | 50.07 s#72 | Sep 12, 2026, 04:45 AM |
78 | Qwen3.8 2.4T A95Bqwen3-8-2-4t-a95b | 40.3 tok/s#78 | 1.88 s#51 | 51.46 s#73 | Sep 12, 2026, 04:45 AM |
79 | Kimi K3kimi-k3 | 37.2 tok/s#79 | 2.99 s#57 | 56.76 s#76 | Sep 12, 2026, 04:45 AM |
80 | Qwen3.5qwen3-5 | 22.3 tok/s#80 | 0.45 s#8 | 90.08 s#78 | Sep 12, 2026, 04:45 AM |
81 | Phi 4 Multimodal Instructphi-4-multimodal-instruct | 17.2 tok/s#81 | 0.41 s#5 | 0.41 s#3 | Sep 12, 2026, 04:45 AM |
