| 1 | MoonshotAI: Kimi K2.5 | MoonshotAI | open-source | 1450 | 262K tokens | — |
| 2 | MoonshotAI: Kimi K2.6 | MoonshotAI | open-source | 1430 | 262K tokens | — |
| 3 | Qwen: Qwen3 235B A22B Instruct 2507 | Alibaba | open-source | 1423 | 262K tokens | — |
| 4 | DeepSeek: DeepSeek V3.2 Exp | DeepSeek | open-source | 1423 | 164K tokens | — |
| 5 | DeepSeek: DeepSeek V3.2 | DeepSeek | open-source | 1422 | 131K tokens | — |
| 6 | MoonshotAI: Kimi K2 0905 | MoonshotAI | open-source | 1418 | 262K tokens | — |
| 7 | DeepSeek: DeepSeek V3.1 | DeepSeek | open-source | 1417 | 33K tokens | — |
| 8 | MoonshotAI: Kimi K2 0711 | MoonshotAI | open-source | 1417 | 131K tokens | — |
| 9 | Qwen: Qwen3 VL 235B A22B Instruct | Alibaba | open-source | 1416 | 262K tokens | — |
| 10 | DeepSeek: DeepSeek V3.1 Terminus | DeepSeek | open-source | 1416 | 164K tokens | — |
| 11 | Mistral Large | Mistral AI | open-source | 1415 | 128K tokens | — |
| 12 | Mistral: Mistral Medium 3 | Mistral AI | open-source | 1410 | 131K tokens | 47 tok/s |
| 13 | MiniMax: MiniMax M2.7 | MiniMax | open-source | 1403 | 197K tokens | — |
| 14 | MiniMax: MiniMax M2.5 | MiniMax | open-source | 1400 | 197K tokens | — |
| 15 | DeepSeek: R1 | DeepSeek | open-source | 1398 | 64K tokens | — |
| 16 | AI21: Jamba Large 1.7 | AI21 Labs | open-source | — | 256K tokens | — |
| 17 | AlfredPros: CodeLLaMa 7B Instruct Solidity | AlfredPros | open-source | — | 4K tokens | — |
| 18 | Qwen: Qwen2.5 7B Instruct | Alibaba | open-source | — | 33K tokens | — |
| 19 | Qwen: Qwen2.5 VL 72B Instruct | Alibaba | open-source | — | 32K tokens | — |
| 20 | Qwen: Qwen3 235B A22B Thinking 2507 | Alibaba | open-source | — | 131K tokens | — |
| 21 | Qwen: Qwen3 30B A3B Instruct 2507 | Alibaba | open-source | — | 262K tokens | — |
| 22 | Qwen: Qwen3 30B A3B Thinking 2507 | Alibaba | open-source | — | 131K tokens | — |
| 23 | Qwen: Qwen3 Coder 30B A3B Instruct | Alibaba | open-source | — | 160K tokens | — |
| 24 | Qwen: Qwen3 Next 80B A3B Instruct | Alibaba | open-source | — | 262K tokens | — |
| 25 | Qwen: Qwen3 VL 30B A3B Instruct | Alibaba | open-source | — | 131K tokens | — |
| 26 | Qwen: Qwen3 VL 32B Instruct | Alibaba | open-source | — | 131K tokens | — |
| 27 | Qwen: Qwen3 VL 8B Instruct | Alibaba | open-source | — | 131K tokens | — |
| 28 | Qwen2.5 72B Instruct | Alibaba | open-source | — | 33K tokens | — |
| 29 | Qwen2.5 Coder 32B Instruct | Alibaba | open-source | — | 33K tokens | — |
| 30 | AllenAI: Olmo 3 32B Think | AllenAI | open-source | — | 66K tokens | — |
| 31 | AllenAI: Olmo 3.1 32B Instruct | AllenAI | open-source | — | 66K tokens | — |
| 32 | Arcee AI: Trinity Large Thinking | Arcee AI | open-source | — | 262K tokens | — |
| 33 | Arcee AI: Trinity Mini | Arcee AI | open-source | — | 131K tokens | — |
| 34 | Baidu: ERNIE 4.5 21B A3B Thinking | Baidu | open-source | — | 131K tokens | — |
| 35 | Baidu: ERNIE 4.5 300B A47B | Baidu | open-source | — | 123K tokens | — |
| 36 | Baidu: ERNIE 4.5 VL 28B A3B | Baidu | open-source | — | 30K tokens | — |
| 37 | Baidu: ERNIE 4.5 VL 424B A47B | Baidu | open-source | — | 123K tokens | — |
| 38 | ByteDance: UI-TARS 7B | ByteDance | open-source | — | 128K tokens | — |
| 39 | DeepSeek: DeepSeek V3 | DeepSeek | open-source | — | 164K tokens | — |
| 40 | DeepSeek: DeepSeek V3.2 Speciale | DeepSeek | open-source | — | 164K tokens | — |
| 41 | DeepSeek: DeepSeek V4 Flash | DeepSeek | open-source | — | 1.0M tokens | 74 tok/s |
| 42 | DeepSeek: DeepSeek V4 Pro | DeepSeek | open-source | — | 1.0M tokens | — |
| 43 | DeepSeek: R1 Distill Llama 70B | DeepSeek | open-source | — | 131K tokens | 44 tok/s |
| 44 | DeepSeek: R1 Distill Qwen 32B | DeepSeek | open-source | — | 33K tokens | — |
| 45 | EssentialAI: Rnj 1 Instruct | EssentialAI | open-source | — | 33K tokens | — |
| 46 | Goliath 120B | Goliath 120B | open-source | — | 6K tokens | — |
| 47 | Google: Lyria 3 Clip Preview | Google | API grátis | — | 1.0M tokens | — |
| 48 | Google: Lyria 3 Pro Preview | Google | API grátis | — | 1.0M tokens | — |
| 49 | IBM: Granite 4.0 Micro | IBM | open-source | — | 131K tokens | — |
| 50 | LiquidAI: LFM2-24B-A2B | LiquidAI | open-source | — | 33K tokens | 105 tok/s |
| 51 | Magnum v4 72B | Magnum v4 72B | open-source | — | 16K tokens | — |
| 52 | Llama Guard 3 8B | Meta | open-source | — | 131K tokens | — |
| 53 | Meta: Llama 3 70B Instruct | Meta | open-source | — | 8K tokens | — |
| 54 | Meta: Llama 3 8B Instruct | Meta | open-source | — | 8K tokens | — |
| 55 | Meta: Llama 3.1 70B Instruct | Meta | open-source | — | 131K tokens | — |
| 56 | Meta: Llama 3.1 8B Instruct | Meta | open-source | — | 16K tokens | — |
| 57 | Meta: Llama 3.2 11B Vision Instruct | Meta | open-source | — | 131K tokens | — |
| 58 | Meta: Llama 3.2 1B Instruct | Meta | open-source | — | 60K tokens | — |
| 59 | Meta: Llama 3.2 3B Instruct | Meta | open-source | — | 80K tokens | — |
| 60 | Meta: Llama 3.3 70B Instruct | Meta | open-source | — | 131K tokens | — |
| 61 | Meta: Llama 4 Maverick | Meta | open-source | — | 1.0M tokens | 110 tok/s |
| 62 | Meta: Llama 4 Scout | Meta | open-source | — | 328K tokens | 135 tok/s |
| 63 | Meta: Llama Guard 4 12B | Meta | open-source | — | 164K tokens | — |
| 64 | Microsoft: Phi 4 | Microsoft | open-source | — | 16K tokens | — |
| 65 | WizardLM-2 8x22B | Microsoft | open-source | — | 66K tokens | — |
| 66 | MiniMax: MiniMax M2.1 | MiniMax | open-source | — | 197K tokens | — |
| 67 | MiniMax: MiniMax-01 | MiniMax | open-source | — | 1.0M tokens | — |
| 68 | Mistral: Devstral 2 2512 | Mistral AI | open-source | — | 262K tokens | — |
| 69 | Mistral: Devstral Medium | Mistral AI | open-source | — | 131K tokens | 78 tok/s |
| 70 | Mistral: Devstral Small 1.1 | Mistral AI | open-source | — | 131K tokens | — |
| 71 | Mistral: Ministral 3 14B 2512 | Mistral AI | open-source | — | 262K tokens | — |
| 72 | Mistral: Ministral 3 3B 2512 | Mistral AI | open-source | — | 131K tokens | — |
| 73 | Mistral: Ministral 3 8B 2512 | Mistral AI | open-source | — | 262K tokens | — |
| 74 | Mistral: Mistral 7B Instruct v0.1 | Mistral AI | open-source | — | 3K tokens | — |
| 75 | Mistral: Mistral Medium 3.1 | Mistral AI | open-source | — | 131K tokens | 93 tok/s |
| 76 | Mistral: Mistral Medium 3.5 | Mistral AI | open-source | — | 262K tokens | 161 tok/s |
| 77 | Mistral: Mistral Nemo | Mistral AI | open-source | — | 131K tokens | — |
| 78 | Mistral: Mistral Small 3.1 24B | Mistral AI | open-source | — | 128K tokens | — |
| 79 | Mistral: Mistral Small 3.2 24B | Mistral AI | open-source | — | 128K tokens | — |
| 80 | Mistral: Mistral Small 4 | Mistral AI | open-source | — | 262K tokens | 169 tok/s |
| 81 | Mistral: Mistral Small Creative | Mistral AI | open-source | — | 33K tokens | — |
| 82 | Mistral: Mixtral 8x22B Instruct | Mistral AI | open-source | — | 66K tokens | — |
| 83 | Mistral: Mixtral 8x7B Instruct | Mistral AI | open-source | — | 33K tokens | — |
| 84 | Mistral: Saba | Mistral AI | open-source | — | 33K tokens | — |
| 85 | Mistral: Voxtral Small 24B 2507 | Mistral AI | open-source | — | 32K tokens | — |
| 86 | MythoMax 13B | MythoMax 13B | open-source | — | 4K tokens | — |
| 87 | Nex AGI: DeepSeek V3.1 Nex N1 | Nex AGI | open-source | — | 131K tokens | — |
| 88 | Nous: Hermes 3 405B Instruct | Nous | open-source | — | 131K tokens | — |
| 89 | Nous: Hermes 3 70B Instruct | Nous | open-source | — | 131K tokens | — |
| 90 | Nous: Hermes 4 405B | Nous | open-source | — | 131K tokens | — |
| 91 | Nous: Hermes 4 70B | Nous | open-source | — | 131K tokens | — |
| 92 | NousResearch: Hermes 2 Pro - Llama-3 8B | NousResearch | open-source | — | 8K tokens | — |
| 93 | NVIDIA: Llama 3.1 Nemotron 70B Instruct | NVIDIA | open-source | — | 131K tokens | — |