| 模型 | 上下文 | 输入价 | 输出价 | 吞吐 | 延迟 | 服务商 | Agents | 编程 | 推理 | 知识 | 数学 | 多语言 | 多模态 | 指令遵循 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| BGE Large ZH V1.5 | — | — | — | — | — | +22 | — | — | — | — | — | — | — | — |
| Gemini 1.5 Flash | — | — | — | 200 t/s | 1.17s | +17 | — | 27 | 39.3 | — | 50.4 | — | — | — |
| Claude 3 Opus | — | — | — | — | — | +11 | — | 23.5 | 40.5 | 21.3 | 33.7 | — | — | — |
| FLUX Dev | — | — | — | — | — | +16 | — | — | — | — | — | — | — | — |
| Gemini 3.0 Flash | — | — | — | 28 t/s | 20.33s | +4 | — | — | — | — | — | — | — | — |
| Gemini 1.5 Flash 002 | — | — | — | 155 t/s | 1.62s | +1 | — | — | — | — | — | — | — | — |
| Ring Flash 2.0 | — | — | — | 107 t/s | 6.09s | +5 | — | 39.8 | 53.6 | — | — | — | — | — |
| DeepSeek R1 MetaSearch | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Qwen3.0 | — | — | — | 165 t/s | 1.77s | +12 | — | — | — | — | — | — | — | — |
| MiniMax Hailuo 02 | — | — | — | — | — | +11 | — | — | — | — | — | — | — | — |
| Gemini 1.0 Pro Vision | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FLUX.2 Max | — | — | — | — | — | +8 | — | — | — | — | — | — | — | — |
| GLM-4.5 Flash | — | — | — | 29 t/s | 17.14s | +43 | — | — | — | — | — | — | — | — |
| Gemini Embedding 2 | — | — | — | — | — | +30 | — | — | — | — | — | — | — | — |
| GLM-Z1 Air | — | — | — | 53 t/s | 0.30s | +8 | — | — | — | — | — | — | — | — |
| Jamba 1.5 Large Instruct | — | — | — | 56 t/s | 0.29s | +17 | — | — | — | — | — | — | — | — |
| Step3 | — | — | — | — | — | +2 | — | — | — | — | — | — | — | — |
| BGE Large EN V1.5 | — | — | — | — | — | +17 | — | — | — | — | — | — | — | — |
| GLM-4.7 FlashX | — | — | — | 45 t/s | 22.27s | +2 | — | — | — | — | — | — | — | — |
| Phi 4 Multimodal Instruct | — | — | — | 83 t/s | 0.39s | +29 | — | 12.1 | 28.1 | — | 39.3 | — | — | — |
| Llama 3.2 Nemoretriever 300m Embed v1 | — | — | — | — | — | +14 | — | — | — | — | — | — | — | — |
| GLM-5.1 | 202.8K | — | — | 48 t/s | 14.18s | +166 | 62.4 | 47.7 | 45.4 | 42.4 | 88.9 | — | 100 | 76.3 |
| TeleSpeechASR | — | — | — | — | — | +11 | — | — | — | — | — | — | — | — |
| GLM-4.6V Flash | — | — | — | 111 t/s | 11.08s | +29 | — | — | — | — | — | — | — | — |
| GLM-4 Plus | — | — | — | — | — | +14 | — | — | — | — | — | — | — | — |
| PanGu Pro MoE | — | — | — | — | — | +4 | — | — | — | — | — | — | — | — |
| Solar Instruct | — | — | — | — | — | +2 | — | — | — | — | — | — | — | — |
| GLM-4 Air | — | — | — | 66 t/s | 0.33s | +24 | — | — | — | — | — | — | — | — |
| Gemini Imagen | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepSeek Coder Instruct | — | — | — | — | — | +4 | — | — | — | — | — | — | — | — |
| GLM-4 Flash | — | — | — | 31 t/s | 0.94s | +50 | — | — | — | — | — | — | — | — |
| Kimi K2 Turbo | — | — | — | 83 t/s | 3.73s | +9 | — | — | — | — | — | — | — | — |
| MiniMax M2.1 HighSpeed | — | — | — | — | — | +17 | — | — | — | — | — | — | — | — |
| Italia Instruct | — | — | — | 36 t/s | 0.49s | +3 | — | — | — | — | — | — | — | — |
| Claude Opus 4.6 Max | — | — | — | — | — | +20 | — | — | — | — | — | — | — | — |
| GLM-4.5-X | — | — | — | 69 t/s | 12.39s | +22 | — | — | — | — | — | — | — | — |
| Phi 3.5 MoE Instruct | — | — | — | — | — | +20 | — | — | — | — | — | — | — | — |
| Qwen3.5 Plus Thinking | — | — | — | — | — | +7 | — | — | — | — | — | — | — | — |
| Qwen3.5 Max | — | — | — | 31 t/s | 29.33s | +7 | — | — | — | — | — | — | — | — |
| GLM-4.1v Thinking FlashX | — | — | — | — | — | +12 | — | — | — | — | — | — | — | — |
| Arctic Embed L | — | — | — | — | — | +18 | — | — | — | — | — | — | — | — |
| Colosseum Instruct | — | — | — | — | — | +1 | — | — | — | — | — | — | — | — |
| Gemini Live 2.5 Flash | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GLM-4V Flash | — | — | — | 56 t/s | 0.62s | +14 | — | — | — | — | — | — | — | — |
| Gemini Pro Vision | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepSeek Prover v2 | — | — | — | — | — | +6 | — | — | — | — | — | — | — | — |
| MiniMax M2.5 HighSpeed | — | — | — | 56 t/s | 5.39s | +31 | — | — | — | — | — | — | — | — |
| GPT-3.5 Net | — | — | — | — | — | +1 | — | — | — | — | — | — | — | — |
| MiMo-V2-Omni | — | — | — | 83 t/s | 3.43s | +57 | 68.2 | 46.5 | 43.4 | 40.2 | — | — | 69.9 | 53.5 |
| Gemini 2.0 Flash Live 001 | — | — | — | — | — | +1 | — | — | — | — | — | — | — | — |
| GPT-4o Transcribe | — | — | — | — | — | +35 | — | — | — | — | — | — | — | — |
| Gemini Embedding | — | — | — | — | — | +8 | — | — | — | — | — | — | — | — |
| GLM-4 AirX | — | — | — | 95 t/s | 3.81s | +18 | — | — | — | — | — | — | — | — |
| DeepSeek Coder v2 Instruct | — | — | — | — | — | +2 | — | — | — | — | — | — | — | — |
| Gemini 3.1 Flash | — | — | — | — | — | +15 | — | — | — | — | — | — | — | — |
| Qwen3.6 Plus Thinking | — | — | — | — | — | +7 | — | — | — | — | — | — | — | — |
| Gemini 2.5 Pro 1M | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MiMo-V2-Pro | — | — | — | 49 t/s | 3.35s | +66 | 51.2 | 51 | 44.1 | 36.2 | — | — | — | 68.8 |
| GLM-4.1v Thinking Flash | — | — | — | 85 t/s | 7.49s | +13 | — | — | — | — | — | — | — | — |
| GPT-4o DALL-E | — | — | — | — | — | — | — | — | — | — | — | — | — |