AI模型性能排行榜(从夯到拉天梯榜)
基于全球权威 arena.ai(LMSYS Chatbot Arena)盲测数据 · 中小学生都能看懂的5档战力天梯图
Agent 智能体总榜 · 总指挥综合实测 (Agent Arena)
考核大模型作为“总指挥”自主调用工具、敲代码、查资料并搞定复杂任务的实操能力(收录 46 款主流模型)
Agent 智能体总榜 · 详细排行榜 (TOP 50)
实测盲测 Elo 胜率综合算法排序| 排名 | 模型名称 | 研发机构 | Elo评分 / 战力 | 链接 |
|---|---|---|---|---|
| 🥇 1 | Claude Fable 5.1 (Max) | Anthropic | 官网 ↗ | |
| 🥈 2 | GPT 6 Astra (Max) | OpenAI | 官网 ↗ | |
| 🥉 3 | Claude Opus 5 (High) | Anthropic | 官网 ↗ | |
| 4 | Claude Opus 5 (Max) | Anthropic | 官网 ↗ | |
| 5 | Claude Fable 5 (High) | Anthropic | 官网 ↗ | |
| 6 | Claude Opus 4.8 (High) | Anthropic | 官网 ↗ | |
| 7 | GPT 5.6 Sol (xHigh) | OpenAI | 官网 ↗ | |
| 8 | Kimi K3 (Max) | Moonshot | 官网 ↗ | |
| 9 | Claude Sonnet 5 (High) | Anthropic | 官网 ↗ | |
| 10 | GPT 5.5 (xHigh) | OpenAI | 官网 ↗ | |
| 11 | Hy4 preview | Tencent | 官网 ↗ | |
| 12 | Deepseek V4.1 Flash (Max) | DeepSeek | 官网 ↗ | |
| 13 | Gemini 3.8 Flash (High) | 官网 ↗ | ||
| 14 | GLM 5.2 (Max) | Z.ai | 官网 ↗ | |
| 15 | Muse Spark 1.3 (Max) | Meta | 官网 ↗ | |
| 16 | DeepSeek V4 Pro (High) (0813) | DeepSeek | 官网 ↗ | |
| 17 | Qwen3.8 Max | Alibaba | 官网 ↗ | |
| 18 | GLM 5.3 (Max) | Z.ai | 官网 ↗ | |
| 19 | Grok 4.5 | SpaceXAI | 官网 ↗ | |
| 20 | GPT 5.5 | OpenAI | 官网 ↗ | |
| 21 | Grok 4.6 (xHigh) | SpaceXAI | 官网 ↗ | |
| 22 | Deepseek V4 Flash (High) (20260731) | DeepSeek | 官网 ↗ | |
| 23 | GPT 5.6 Terra (xHigh) | OpenAI | 官网 ↗ | |
| 24 | GPT 5.4 (High) | OpenAI | 官网 ↗ | |
| 25 | GLM 5.3 Flash | Z.ai | 官网 ↗ | |
| 26 | Qwen3.8 Flash Next | Alibaba | 官网 ↗ | |
| 27 | GPT 5.6 Luna (xHigh) | OpenAI | 官网 ↗ | |
| 28 | Gemini 3.7 Flash (High) | 官网 ↗ | ||
| 29 | Qwen 3.8 27B | Alibaba | 官网 ↗ | |
| 30 | DeepSeek V4 Pro | DeepSeek | 官网 ↗ | |
| 31 | Claude Sonnet 4.6 | Anthropic | 官网 ↗ | |
| 32 | Muse Spark 1.2 (xHigh) | Meta | 官网 ↗ | |
| 33 | Muse Spark 1.1 | Meta | 官网 ↗ | |
| 34 | Qwen3.7 Max | Alibaba | - | |
| 35 | Hy3 | Tencent | 官网 ↗ | |
| 36 | Mimo V2.5 Pro | Xiaomi | 官网 ↗ | |
| 37 | Minimax M3 | MiniMax | 官网 ↗ | |
| 38 | Gemini 3.1 Pro Preview | 官网 ↗ | ||
| 39 | Gemini 3.6 Flash (High) | 官网 ↗ | ||
| 40 | Qwen3.7 Plus | Alibaba | 官网 ↗ | |
| 41 | Inkling Small | Thinky | 官网 ↗ | |
| 42 | Inkling | Thinky | 官网 ↗ | |
| 43 | Mistral Medium 3.5 | Mistral | 官网 ↗ | |
| 44 | Minimax M2.7 | MiniMax | 官网 ↗ | |
| 45 | Gemini 3.5 Flash Lite | 官网 ↗ | ||
| 46 | Solar Pro 4 | Upstage | 官网 ↗ |
文本能力榜
日常对话与综合语言 (Text Arena) · 考核大模型的日常对话、中英写作、知识答疑与逻辑推理综合能力
文本能力榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | claude-fable-5-high | Anthropic | ↗ | |
| 🥈 2 | claude-opus-4-6-high | Anthropic | ↗ | |
| 🥉 3 | claude-opus-4-7-high | Anthropic | ↗ | |
| 4 | muse-spark-1.2 (xHigh) | Meta | ↗ | |
| 5 | claude-fable-5.1-max | Anthropic | ↗ | |
| 6 | claude-opus-4-6 | Anthropic | ↗ | |
| 7 | claude-opus-4-7 | Anthropic | ↗ | |
| 8 | muse-spark-1.3-max | Meta | ↗ | |
| 9 | gemini-3.8-flash-high | ↗ | ||
| 10 | claude-opus-5-high | Anthropic | ↗ | |
| 11 | muse-spark-1.1 | Meta | ↗ | |
| 12 | gemini-3.7-flash-high | ↗ | ||
| 13 | muse-spark | Meta | ↗ | |
| 14 | claude-opus-5-max | Anthropic | ↗ | |
| 15 | gemini-3.1-pro-preview | ↗ | ||
| 16 | gemini-3-pro | ↗ | ||
| 17 | kimi-k3-max | Moonshot | ↗ | |
| 18 | gpt-5.6-sol-xhigh | OpenAI | ↗ | |
| 19 | glm-5.3-max | Z.ai | ↗ | |
| 20 | gpt-5.5-high | OpenAI | ↗ | |
| 21 | claude-opus-4-8-high | Anthropic | ↗ | |
| 22 | qwen3.8-max | Alibaba | ↗ | |
| 23 | gemini-3.6-flash-high | ↗ | ||
| 24 | gpt-6-astra-max | OpenAI | ↗ | |
| 25 | gemini-3.5-flash-high | ↗ | ||
| 26 | gpt-5.4-high | OpenAI | ↗ | |
| 27 | gpt-5.2-chat-latest-20260210 | OpenAI | ↗ | |
| 28 | gpt-5.5 | OpenAI | ↗ | |
| 29 | glm-5.3-flash | Z.ai | ↗ | |
| 30 | grok-4.20-beta1 | SpaceXAI | ↗ | |
| 31 | gemini-3.5-flash-medium | ↗ | ||
| 32 | gpt-5.5-instant | OpenAI | ↗ | |
| 33 | gemini-3-flash | ↗ | ||
| 34 | qwen3.7-max-preview | Alibaba | ↗ | |
| 35 | claude-opus-4-8 | Anthropic | ↗ | |
| 36 | claude-opus-4-5-20251101-high-32k | Anthropic | ↗ | |
| 37 | claude-sonnet-4-6 | Anthropic | ↗ | |
| 38 | glm-5.2-max | Z.ai | ↗ | |
| 39 | grok-4.20-beta-0309-reasoning | SpaceXAI | ↗ | |
| 40 | grok-4.20-multi-agent-beta-0309 | SpaceXAI | ↗ | |
| 41 | claude-opus-4-5-20251101 | Anthropic | ↗ | |
| 42 | grok-4.5 | SpaceXAI | ↗ | |
| 43 | ernie-5.1 | Baidu | ↗ | |
| 44 | mimo-v2.5-pro | Xiaomi | ↗ | |
| 45 | gpt-5.6-terra-xhigh | OpenAI | ↗ | |
| 46 | gpt-5.4 | OpenAI | ↗ | |
| 47 | glm-5.1 | Z.ai | ↗ | |
| 48 | grok-4.1-thinking | SpaceXAI | ↗ | |
| 49 | qwen3.5-max-preview | Alibaba | ↗ | |
| 50 | deepseek-v4-pro-high-20260813 | DeepSeek | ↗ |
网页开发与代码榜
全栈开发与代码编写 (WebDev Arena) · 考核大模型写前端网页、全栈代码、修Bug与解决真实编程问题的能力
网页开发与代码榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | gpt-6-astra-max | OpenAI | ↗ | |
| 🥈 2 | claude-fable-5.1-max | Anthropic | ↗ | |
| 🥉 3 | claude-opus-5-max | Anthropic | ↗ | |
| 4 | qwen3.8-max-0902 | Alibaba | ↗ | |
| 5 | kimi-k3-max | Moonshot | ↗ | |
| 6 | qwen3.8-max | Alibaba | ↗ | |
| 7 | claude-opus-5-high | Anthropic | ↗ | |
| 8 | muse-spark-1.3-max | Meta | ↗ | |
| 9 | qwen3.8-flash-next | Alibaba | ↗ | |
| 10 | claude-fable-5-high | Anthropic | ↗ | |
| 11 | hy4-preview | Tencent | ↗ | |
| 12 | muse-spark-1.3 (xHigh) | Meta | ↗ | |
| 13 | grok-4.6-high | SpaceXAI | ↗ | |
| 14 | gpt-5.6-sol-xhigh (codex-harness) | OpenAI | ↗ | |
| 15 | glm-5.3-max | Z.ai | ↗ | |
| 16 | deepseek-v4.1-flash-max | DeepSeek | ↗ | |
| 17 | glm-5.3-flash | Z.ai | ↗ | |
| 18 | qwen3.8-27b | Alibaba | ↗ | |
| 19 | glm-5.2-max | Z.ai | ↗ | |
| 20 | gemini-3.7-flash-high | ↗ | ||
| 21 | deepseek-v4-pro-high-20260813 | DeepSeek | ↗ | |
| 22 | deepseek-v4-flash-high | DeepSeek | ↗ | |
| 23 | gemini-3.8-flash-high | ↗ | ||
| 24 | claude-opus-4-8-high | Anthropic | ↗ | |
| 25 | claude-opus-4-7 | Anthropic | ↗ | |
| 26 | claude-opus-4-7-high | Anthropic | ↗ | |
| 27 | grok-4.5 | SpaceXAI | ↗ | |
| 28 | claude-opus-4-6-high | Anthropic | ↗ | |
| 29 | muse-spark-1.1 | Meta | ↗ | |
| 30 | claude-opus-4-8 | Anthropic | ↗ | |
| 31 | claude-opus-4-6 | Anthropic | ↗ | |
| 32 | gemini-3.6-flash-high | ↗ | ||
| 33 | claude-sonnet-5-high | Anthropic | ↗ | |
| 34 | muse-spark-1.2 (xHigh) | Meta | ↗ | |
| 35 | gpt-5.6-terra-xhigh (codex-harness) | OpenAI | ↗ | |
| 36 | claude-sonnet-4-6 | Anthropic | ↗ | |
| 37 | gpt-5.6-luna-xhigh (codex-harness) | OpenAI | ↗ | |
| 38 | seed-2.1-pro-preview | Bytedance | ↗ | |
| 39 | qwen3.7-max-20260517 | Alibaba | ↗ | |
| 40 | hy3 | Tencent | ↗ | |
| 41 | gpt-5.5-xhigh (codex-harness) | OpenAI | ↗ | |
| 42 | kimi-k2.6 | Moonshot | ↗ | |
| 43 | glm-5.1 | Z.ai | ↗ | |
| 44 | gemini-3.5-flash-high | ↗ | ||
| 45 | claude-opus-4-5-20251101-high-32k | Anthropic | ↗ | |
| 46 | gemini-3.5-flash-medium | ↗ | ||
| 47 | gpt-5.5-high (codex-harness) | OpenAI | ↗ | |
| 48 | minimax-m3 | MiniMax | ↗ | |
| 49 | qwen3.6-max-preview | Alibaba | ↗ | |
| 50 | mimo-v2.5-pro | Xiaomi | ↗ |
视觉看图识图榜
读图看表火眼金睛 (Vision Arena) · 考核大模型的“眼睛”,看懂复杂照片、图表、界面截图并深入推理解析的能力
视觉看图识图榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | claude-fable-5-high | Anthropic | ↗ | |
| 🥈 2 | qwen3.8-max | Alibaba | ↗ | |
| 🥉 3 | claude-opus-4-7-high | Anthropic | ↗ | |
| 4 | claude-opus-4-7 | Anthropic | ↗ | |
| 5 | claude-opus-4-6-high | Anthropic | ↗ | |
| 6 | muse-spark-1.3-max | Meta | ↗ | |
| 7 | muse-spark | Meta | ↗ | |
| 8 | claude-opus-4-6 | Anthropic | ↗ | |
| 9 | muse-spark-1.2 (xHigh) | Meta | ↗ | |
| 10 | claude-opus-5-high | Anthropic | ↗ | |
| 11 | claude-fable-5.1-max | Anthropic | ↗ | |
| 12 | gemini-3-pro | ↗ | ||
| 13 | gpt-5.5 | OpenAI | ↗ | |
| 14 | gpt-5.6-sol-xhigh | OpenAI | ↗ | |
| 15 | gpt-5.4-high | OpenAI | ↗ | |
| 16 | gpt-6-astra-max | OpenAI | ↗ | |
| 17 | gemini-3.5-flash-high | ↗ | ||
| 18 | claude-opus-4-8-high | Anthropic | ↗ | |
| 19 | gpt-5.5-high | OpenAI | ↗ | |
| 20 | gemini-3.5-flash-medium | ↗ | ||
| 21 | gemini-3.6-flash-high | ↗ | ||
| 22 | muse-spark-1.1 | Meta | ↗ | |
| 23 | gpt-5.4 | OpenAI | ↗ | |
| 24 | grok-4.5 | SpaceXAI | ↗ | |
| 25 | claude-opus-4-8 | Anthropic | ↗ | |
| 26 | gemini-3.1-pro-preview | ↗ | ||
| 27 | gpt-5.2-chat-latest-20260210 | OpenAI | ↗ | |
| 28 | gpt-5.5-instant | OpenAI | ↗ | |
| 29 | claude-sonnet-4-6 | Anthropic | ↗ | |
| 30 | glm-5.3-flash | Z.ai | ↗ | |
| 31 | gemini-3-flash | ↗ | ||
| 32 | gpt-5.6-terra-xhigh | OpenAI | ↗ | |
| 33 | claude-sonnet-5-high | Anthropic | ↗ | |
| 34 | gemini-3.5-flash-lite | ↗ | ||
| 35 | grok-4.6-high | SpaceXAI | ↗ | |
| 36 | qwen3.7-plus | Alibaba | ↗ | |
| 37 | kimi-k2.6 | Moonshot | ↗ | |
| 38 | gemma-4-31b | ↗ | ||
| 39 | gemini-3-flash (thinking-minimal) | ↗ | ||
| 40 | gpt-5.6-luna-xhigh | OpenAI | ↗ | |
| 41 | dola-seed-2.0-pro | Bytedance | ↗ | |
| 42 | grok-4.20-beta-0309-reasoning | SpaceXAI | ↗ | |
| 43 | gpt-5.4-mini-high | OpenAI | ↗ | |
| 44 | grok-4.20-multi-agent-beta-0309 | SpaceXAI | ↗ | |
| 45 | gpt-5.1-high | OpenAI | ↗ | |
| 46 | kimi-k2.5-thinking | Moonshot | ↗ | |
| 47 | qwen3.5-397b-a17b | Alibaba | ↗ | |
| 48 | gemini-2.5-pro | ↗ | ||
| 49 | qwen3.8-27b | Alibaba | ↗ | |
| 50 | gpt-5.2-high | OpenAI | ↗ |
长文档分析提炼榜
超长万字速读记忆 (Document Arena) · 考核超长文档阅读记忆力,快速速读万字长文、专业论文与长财报并精准提炼总结
长文档分析提炼榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | claude-opus-5-high | Anthropic | ↗ | |
| 🥈 2 | claude-fable-5.1-max | Anthropic | ↗ | |
| 🥉 3 | claude-opus-4-6-high | Anthropic | ↗ | |
| 4 | claude-opus-4-6 | Anthropic | ↗ | |
| 5 | claude-fable-5-high | Anthropic | ↗ | |
| 6 | claude-opus-4-7 | Anthropic | ↗ | |
| 7 | claude-opus-4-7-high | Anthropic | ↗ | |
| 8 | gpt-5.5 | OpenAI | ↗ | |
| 9 | gpt-5.5-high | OpenAI | ↗ | |
| 10 | gpt-5.6-sol-xhigh | OpenAI | ↗ | |
| 11 | claude-sonnet-4-6 | Anthropic | ↗ | |
| 12 | claude-opus-4-8-high | Anthropic | ↗ | |
| 13 | gpt-5.6-terra-xhigh | OpenAI | ↗ | |
| 14 | gpt-5.4 | OpenAI | ↗ | |
| 15 | muse-spark-1.3-max | Meta | ↗ | |
| 16 | gpt-6-astra-max | OpenAI | ↗ | |
| 17 | claude-sonnet-5-high | Anthropic | ↗ | |
| 18 | muse-spark-1.1 | Meta | ↗ | |
| 19 | claude-opus-4-8 | Anthropic | ↗ | |
| 20 | gemini-3.5-flash-medium | ↗ | ||
| 21 | claude-opus-4-5-20251101 | Anthropic | ↗ | |
| 22 | gemini-3.5-flash-high | ↗ | ||
| 23 | gpt-5.6-luna-xhigh | OpenAI | ↗ | |
| 24 | gemini-3.6-flash-high | ↗ | ||
| 25 | grok-4.6-high | SpaceXAI | ↗ | |
| 26 | grok-4.5 | SpaceXAI | ↗ | |
| 27 | kimi-k2.6 | Moonshot | ↗ | |
| 28 | claude-sonnet-4-5-20250929 | Anthropic | ↗ | |
| 29 | muse-spark | Meta | ↗ | |
| 30 | qwen3.7-plus | Alibaba | ↗ | |
| 31 | gemini-3.1-pro-preview | ↗ | ||
| 32 | minimax-m3 | MiniMax | ↗ | |
| 33 | gemini-3-pro | ↗ | ||
| 34 | kimi-k2.5-thinking | Moonshot | ↗ | |
| 35 | gemma-4-31b | ↗ | ||
| 36 | gemini-2.5-pro | ↗ | ||
| 37 | claude-haiku-4-5-20251001 | Anthropic | ↗ | |
| 38 | grok-4.20-beta-0309-reasoning | SpaceXAI | ↗ | |
| 39 | glm-5v-turbo | Z.ai | ↗ | |
| 40 | gemini-3-flash | ↗ | ||
| 41 | gpt-5.2-high | OpenAI | ↗ | |
| 42 | gpt-5.1 | OpenAI | ↗ | |
| 43 | gpt-5.5-instant | OpenAI | ↗ | |
| 44 | gpt-5.2 | OpenAI | ↗ |
文生图绘画榜
神笔马良一键出图 (Text-to-Image) · 考核输入文字提示词直接生成高清逼真、极富质感的画作与设计素材的能力
文生图绘画榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | gpt-image-2.5-sunburst | OpenAI | ↗ | |
| 🥈 2 | gpt-image-2.5-flare | OpenAI | ↗ | |
| 🥉 3 | gpt-image-2 (medium) | OpenAI | ↗ | |
| 4 | mai-image-2.6 | Microsoft AI | ↗ | |
| 5 | grok-imagine-image-2.0 (low) | SpaceXAI | ↗ | |
| 6 | reve-2.1 | Reve | ↗ | |
| 7 | muse-image | Meta | ↗ | |
| 8 | reve-2.0 | Reve | ↗ | |
| 9 | gemini-3.1-flash-image (nano-banana-2) [web-search] | ↗ | ||
| 10 | seedream-5.0-pro | Bytedance | ↗ | |
| 11 | qwen-image-3.0-pro | Alibaba | ↗ | |
| 12 | mai-image-2.5 | Microsoft AI | ↗ | |
| 13 | gemini-3.1-flash-lite-image (nano-banana-2-lite) | ↗ | ||
| 14 | gemini-3-pro-image-2k (nano-banana-pro) | ↗ | ||
| 15 | gpt-image-1.5-high-fidelity | OpenAI | ↗ | |
| 16 | gemini-3-pro-image-preview (nano-banana-pro) | ↗ | ||
| 17 | ideogram-4.0-quality | Ideogram | ↗ | |
| 18 | qwen-image-2.0-pro-2026-06-22 | Alibaba | ↗ | |
| 19 | uni-1.1-max | Luma AI | ↗ | |
| 20 | mai-image-2 | Microsoft AI | ↗ | |
| 21 | uni-1.1 | Luma AI | ↗ | |
| 22 | grok-imagine-image | SpaceXAI | ↗ | |
| 23 | recraft-v4.1-utility-pro | Recraft | ↗ | |
| 24 | Cosmos3-Super-Text2Image (Agentic) | Nvidia | ↗ | |
| 25 | flux-2-max | Black Forest Labs | ↗ | |
| 26 | grok-imagine-image-pro | SpaceXAI | ↗ | |
| 27 | flux-2-flex | Black Forest Labs | ↗ | |
| 28 | flux-2-pro | Black Forest Labs | ↗ | |
| 29 | reve-v1.5 | Reve | ↗ | |
| 30 | Cosmos3-Super-Text2Image | Nvidia | ↗ | |
| 31 | hunyuan-image-3.0 | Tencent | ↗ | |
| 32 | gemini-2.5-flash-image-preview (nano-banana) | ↗ | ||
| 33 | imagen-ultra-4.0-generate-001 | ↗ | ||
| 34 | seedream-4.5 | Bytedance | ↗ | |
| 35 | flux-2-dev | Black Forest Labs | ↗ | |
| 36 | seedream-4-2k | Bytedance | ↗ | |
| 37 | seedream-5.0-lite | Bytedance | ↗ | |
| 38 | wan2.6-t2i | Alibaba | ↗ | |
| 39 | recraft-v4.1-pro | Recraft | ↗ | |
| 40 | imagen-4.0-generate-001 | ↗ | ||
| 41 | qwen-image-2512 | Alibaba | ↗ | |
| 42 | krea-2-medium | Krea | ↗ | |
| 43 | wan2.5-t2i-preview | Alibaba | ↗ | |
| 44 | hidream-o1-image | HiDream | ↗ | |
| 45 | seedream-4-fal | Bytedance | ↗ | |
| 46 | gpt-image-1 | OpenAI | ↗ | |
| 47 | recraft-v4 | Recraft | ↗ | |
| 48 | seedream-4-high-res-fal | Bytedance | ↗ | |
| 49 | krea-2-turbo | Krea | ↗ | |
| 50 | gpt-image-1-mini | OpenAI | ↗ |
图像精修编辑榜
修图调色微调大师 (Image Edit) · 考核指令修图能力,按自然语言指令给图片改细节、换背景、抠图与局部微调重绘
图像精修编辑榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | gpt-image-2.5-sunburst | OpenAI | ↗ | |
| 🥈 2 | gpt-image-2.5-flare | OpenAI | ↗ | |
| 🥉 3 | gpt-image-2 (medium) | OpenAI | ↗ | |
| 4 | grok-imagine-image-2.0 (low) | SpaceXAI | ↗ | |
| 5 | mai-image-2.6 | Microsoft AI | ↗ | |
| 6 | muse-image | Meta | ↗ | |
| 7 | mai-image-2.5 | Microsoft AI | ↗ | |
| 8 | seedream-5.0-pro | Bytedance | ↗ | |
| 9 | gemini-3-pro-image-2k (nano-banana-pro) | ↗ | ||
| 10 | grok-imagine-image-quality (20260519) | SpaceXAI | ↗ | |
| 11 | chatgpt-image-latest-high-fidelity (20251216) | OpenAI | ↗ | |
| 12 | gemini-3.1-flash-image (nano-banana-2) [web-search] | ↗ | ||
| 13 | gemini-3-pro-image-preview (nano-banana-pro) | ↗ | ||
| 14 | reve-2.1 | Reve | ↗ | |
| 15 | gpt-image-1.5-high-fidelity | OpenAI | ↗ | |
| 16 | reve-2.0 | Reve | ↗ | |
| 17 | uni-1.1-max | Luma AI | ↗ | |
| 18 | grok-imagine-image | SpaceXAI | ↗ | |
| 19 | gemini-3.1-flash-lite-image (nano-banana-2-lite) | ↗ | ||
| 20 | uni-1.1 | Luma AI | ↗ | |
| 21 | qwen-image-2.0-pro-2026-06-22 | Alibaba | ↗ | |
| 22 | hunyuan-image-3.0-instruct | Tencent | ↗ | |
| 23 | wan2.7-image-pro | Alibaba | ↗ | |
| 24 | seedream-4.5 | Bytedance | ↗ | |
| 25 | wan2.7-image | Alibaba | ↗ | |
| 26 | seedream-5.0-lite | Bytedance | ↗ | |
| 27 | gemini-2.5-flash-image-preview (nano-banana) | ↗ | ||
| 28 | seedream-4-2k | Bytedance | ↗ | |
| 29 | flux-2-max | Black Forest Labs | ↗ | |
| 30 | reve-v1.1 | Reve | ↗ | |
| 31 | kling-image-o1 | KlingAI | ↗ | |
| 32 | flux-2-pro | Black Forest Labs | ↗ | |
| 33 | qwen-image-edit | Alibaba | ↗ | |
| 34 | qwen-image-edit-2511 | Alibaba | ↗ | |
| 35 | reve-v1 | Reve | ↗ | |
| 36 | wan2.6-image | Alibaba | ↗ | |
| 37 | flux-2-dev | Black Forest Labs | ↗ | |
| 38 | flux-2-flex | Black Forest Labs | ↗ | |
| 39 | flux-2-klein-9b | Black Forest Labs | ↗ | |
| 40 | seedream-4-high-res-fal | Bytedance | ↗ | |
| 41 | p-image-edit | AI Lab | ↗ | |
| 42 | seedream-4-fal | Bytedance | ↗ | |
| 43 | reve-v1.1-fast | Reve | ↗ | |
| 44 | reve-edit-fast | Reve | ↗ | |
| 45 | flux-2-klein-4b | Black Forest Labs | ↗ | |
| 46 | wan2.5-i2i-preview | Alibaba | ↗ | |
| 47 | flux-1-kontext-max | Black Forest Labs | ↗ | |
| 48 | flux-1-kontext-pro | Black Forest Labs | ↗ | |
| 49 | flux-1-kontext-dev | Black Forest Labs | ↗ | |
| 50 | seededit-3.0 | Bytedance | ↗ |
看图转网页代码榜
设计稿秒变可运行网页 (Image-to-WebDev) · 考核看设计图直接还原代码的能力,给一张UI设计图就能自动生成完整的前端网页代码
看图转网页代码榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | gpt-6-astra-max | OpenAI | ↗ | |
| 🥈 2 | claude-fable-5.1-max | Anthropic | ↗ | |
| 🥉 3 | claude-opus-5-max | Anthropic | ↗ | |
| 4 | muse-spark-1.3-max | Meta | ↗ | |
| 5 | qwen3.8-max-0902 | Alibaba | ↗ | |
| 6 | claude-fable-5-high | Anthropic | ↗ | |
| 7 | qwen3.8-max | Alibaba | ↗ | |
| 8 | gpt-5.6-sol-xhigh (codex-harness) | OpenAI | ↗ | |
| 9 | grok-4.6-high | SpaceXAI | ↗ | |
| 10 | glm-5.3-flash | Z.ai | ↗ | |
| 11 | qwen3.8-27b | Alibaba | ↗ | |
| 12 | kimi-k3-max | Moonshot | ↗ | |
| 13 | claude-opus-4-7-high | Anthropic | ↗ | |
| 14 | claude-opus-4-7 | Anthropic | ↗ | |
| 15 | grok-4.5 | SpaceXAI | ↗ | |
| 16 | gemini-3.6-flash-high | ↗ | ||
| 17 | claude-sonnet-5-high | Anthropic | ↗ | |
| 18 | claude-opus-4-6 | Anthropic | ↗ | |
| 19 | claude-opus-4-6-high | Anthropic | ↗ | |
| 20 | muse-spark-1.1 | Meta | ↗ | |
| 21 | gpt-5.6-terra-xhigh (codex-harness) | OpenAI | ↗ | |
| 22 | gpt-5.5-xhigh (codex-harness) | OpenAI | ↗ | |
| 23 | claude-sonnet-4-6 | Anthropic | ↗ | |
| 24 | kimi-k2.6 | Moonshot | ↗ | |
| 25 | gpt-5.5-high (codex-harness) | OpenAI | ↗ | |
| 26 | kimi-k2.7-code | Moonshot | ↗ | |
| 27 | seed-2.1-pro-preview | Bytedance | ↗ | |
| 28 | gpt-5.6-luna-xhigh (codex-harness) | OpenAI | ↗ | |
| 29 | gpt-5.5 (codex-harness) | OpenAI | ↗ | |
| 30 | gemini-3.5-flash-medium | ↗ | ||
| 31 | gemini-3.5-flash-high | ↗ | ||
| 32 | gemini-3.1-pro-preview | ↗ | ||
| 33 | minimax-m3 | MiniMax | ↗ | |
| 34 | qwen3.6-plus | Alibaba | ↗ | |
| 35 | gemini-3-flash | ↗ | ||
| 36 | gemini-3.5-flash-lite | ↗ | ||
| 37 | gemini-3-pro | ↗ | ||
| 38 | gpt-5.4 | OpenAI | ↗ | |
| 39 | gpt-5.3-codex (codex-harness) | OpenAI | ↗ | |
| 40 | kimi-k2.5-thinking | Moonshot | ↗ | |
| 41 | gemini-3-flash (thinking-minimal) | ↗ | ||
| 42 | glm-5v-turbo | Z.ai | ↗ | |
| 43 | gpt-5.1-high | OpenAI | ↗ | |
| 44 | kimi-k2.5-instant | Moonshot | ↗ | |
| 45 | Inkling Small | Thinky | ↗ | |
| 46 | grok-4.3 | SpaceXAI | ↗ | |
| 47 | mistral-large-3 | Mistral | ↗ | |
| 48 | gpt-5.1 | OpenAI | ↗ | |
| 49 | gemini-3.1-flash-lite-preview | ↗ | ||
| 50 | gemini-2.5-pro | ↗ |
智能搜索找资料榜
联网检索事实核查 (Search Arena) · 考核全网实时搜索与精准溯源能力,快速搜集多方权威资料并给出真实靠谱解答
智能搜索找资料榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | gpt-5.6-sol-xhigh | OpenAI | ↗ | |
| 🥈 2 | claude-opus-4-6-search | Anthropic | ↗ | |
| 🥉 3 | gpt-5.5-search | OpenAI | ↗ | |
| 4 | claude-opus-4-7 | Anthropic | ↗ | |
| 5 | claude-fable-5-high | Anthropic | ↗ | |
| 6 | ernie-5.1 | Baidu | ↗ | |
| 7 | claude-sonnet-4-6-search | Anthropic | ↗ | |
| 8 | grok-4.5 | SpaceXAI | ↗ | |
| 9 | gemini-3.1-pro-grounding | ↗ | ||
| 10 | gemini-3-pro-grounding | ↗ | ||
| 11 | gpt-5.2-search | OpenAI | ↗ | |
| 12 | claude-opus-4-8 | Anthropic | ↗ | |
| 13 | grok-4.20-multi-agent-beta-0309 | SpaceXAI | ↗ | |
| 14 | gpt-5.1-search | OpenAI | ↗ | |
| 15 | gemini-3-flash-grounding | ↗ | ||
| 16 | gpt-5.4-search | OpenAI | ↗ | |
| 17 | claude-sonnet-5-search | Anthropic | ↗ | |
| 18 | grok-4.20-beta1 | SpaceXAI | ↗ | |
| 19 | claude-opus-4-5-search | Anthropic | ↗ | |
| 20 | gpt-5.2-search-non-reasoning | OpenAI | ↗ | |
| 21 | grok-4-1-fast-search | SpaceXAI | ↗ | |
| 22 | grok-4-fast-search | SpaceXAI | ↗ | |
| 23 | grok-4.3 | SpaceXAI | ↗ | |
| 24 | claude-sonnet-4-5-search | Anthropic | ↗ | |
| 25 | claude-opus-4-1-search | Anthropic | ↗ | |
| 26 | o3-search | OpenAI | ↗ | |
| 27 | gemini-2.5-pro-grounding | ↗ | ||
| 28 | grok-4-search | SpaceXAI | ↗ | |
| 29 | ppl-sonar-reasoning-pro-high | Perplexity AI | ↗ | |
| 30 | gpt-5-search | OpenAI | ↗ | |
| 31 | ppl-sonar-pro-high | Perplexity AI | ↗ | |
| 32 | claude-opus-4-search | Anthropic | ↗ | |
| 33 | diffbot-small-xl | Diffbot | ↗ | |
| 34 | api-gpt-4o-search | OpenAI | ↗ |
文生视频榜
一句话拍出高清大片 (Text-to-Video) · 考核AI视频导演水平,输入一句话提示词生成连贯流畅、动作逼真且具电影质感的视频
文生视频榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | gemini-omni-1.1-flash | ↗ | ||
| 🥈 2 | gemini-omni-flash | ↗ | ||
| 🥉 3 | wan3.0 | Alibaba | ↗ | |
| 4 | flux-3-video | Black Forest Labs | ↗ | |
| 5 | grok-imagine-video-1.5-agent | SpaceXAI | ↗ | |
| 6 | dreamina-seedance-2.5-720p | Bytedance | ↗ | |
| 7 | dreamina-seedance-2.0-720p | Bytedance | ↗ | |
| 8 | minimax-h3 | MiniMax | ↗ | |
| 9 | muse-video | Meta | ↗ | |
| 10 | happyhorse-1.0 | Alibaba-ATH | ↗ | |
| 11 | sora-2-pro | OpenAI | ↗ | |
| 12 | veo-3.1-audio | ↗ | ||
| 13 | veo-3.1-audio-1080p | ↗ | ||
| 14 | veo-3.1-fast-audio | ↗ | ||
| 15 | veo-3.1-fast-audio-1080p | ↗ | ||
| 16 | veo-3-fast-audio | ↗ | ||
| 17 | grok-imagine-video-720p | SpaceXAI | ↗ | |
| 18 | sora-2 | OpenAI | ↗ | |
| 19 | wan2.7-t2v | Alibaba | ↗ | |
| 20 | veo-3-audio | ↗ | ||
| 21 | wan2.6-t2v | Alibaba | ↗ | |
| 22 | seedance-v1.5-pro | Bytedance | ↗ | |
| 23 | veo-3 | ↗ | ||
| 24 | veo-3-fast | ↗ | ||
| 25 | wan2.5-t2v-preview | Alibaba | ↗ | |
| 26 | pixverse-v5.6 | AI Lab | ↗ | |
| 27 | runway-gen-4.5 | Runway | ↗ | |
| 28 | kling-2.5-turbo-1080p | KlingAI | ↗ | |
| 29 | kling-2.6-pro | KlingAI | ↗ | |
| 30 | p-video | AI Lab | ↗ | |
| 31 | ray-3 | Luma AI | ↗ | |
| 32 | hailuo-2.3 | MiniMax | ↗ | |
| 33 | kling-o1-pro | KlingAI | ↗ | |
| 34 | hailuo-02-pro | MiniMax | ↗ | |
| 35 | seedance-v1-pro | Bytedance | ↗ | |
| 36 | hailuo-02-standard | MiniMax | ↗ | |
| 37 | kandinsky-5.0-t2v-pro | Kandinsky | ↗ | |
| 38 | hunyuan-video-1.5 | Tencent | ↗ | |
| 39 | veo-2 | ↗ | ||
| 40 | kling-v2.1-master | KlingAI | ↗ | |
| 41 | ltx-2-19b | AI Lab | ↗ | |
| 42 | wan-v2.2-a14b | Alibaba | ↗ | |
| 43 | kandinsky-5.0-t2v-lite | Kandinsky | ↗ | |
| 44 | seedance-v1-lite | Bytedance | ↗ | |
| 45 | sora | OpenAI | ↗ | |
| 46 | ray2 | Luma AI | ↗ | |
| 47 | pika-v2.2 | Pika | ↗ | |
| 48 | mochi-v1 | Genmo AI | ↗ |
图生视频榜
让静态照片生动动起来 (Image-to-Video) · 考核给一张静态照片注入动态生命力,生成自然流畅、符合真实物理规律的连贯动态视频
图生视频榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | minimax-h3 | MiniMax | ↗ | |
| 🥈 2 | gemini-omni-1.1-flash | ↗ | ||
| 🥉 3 | wan3.0 | Alibaba | ↗ | |
| 4 | dreamina-seedance-2.5-720p | Bytedance | ↗ | |
| 5 | dreamina-seedance-2.0-720p | Bytedance | ↗ | |
| 6 | gemini-omni-flash | ↗ | ||
| 7 | grok-imagine-video-1.5-720p | SpaceXAI | ↗ | |
| 8 | hidream-o1-video-1.0 | HiDream | ↗ | |
| 9 | flux-3-video-20260811 | Black Forest Labs | ↗ | |
| 10 | happyhorse-1.0 | Alibaba-ATH | ↗ | |
| 11 | wan2.7-i2v | Alibaba | ↗ | |
| 12 | grok-imagine-video-720p | SpaceXAI | ↗ | |
| 13 | veo-3.1-audio | ↗ | ||
| 14 | veo-3.1-audio-1080p | ↗ | ||
| 15 | veo-3.1-fast-audio | ↗ | ||
| 16 | grok-imagine-video-480p | SpaceXAI | ↗ | |
| 17 | veo-3.1-fast-audio-1080p | ↗ | ||
| 18 | vidu-q3-pro | Shengshu | ↗ | |
| 19 | kling-v3-pro | KlingAI | ↗ | |
| 20 | veo-3-audio | ↗ | ||
| 21 | veo-3-fast-audio | ↗ | ||
| 22 | wan2.5-i2v-preview | Alibaba | ↗ | |
| 23 | wan2.6-i2v | Alibaba | ↗ | |
| 24 | seedance-v1.5-pro | Bytedance | ↗ | |
| 25 | pixverse-v5.6 | AI Lab | ↗ | |
| 26 | kling-2.6-pro | KlingAI | ↗ | |
| 27 | kling-2.5-turbo-1080p | KlingAI | ↗ | |
| 28 | seedance-v1-pro | Bytedance | ↗ | |
| 29 | hailuo-2.3 | MiniMax | ↗ | |
| 30 | veo-3 | ↗ | ||
| 31 | veo-3-fast | ↗ | ||
| 32 | p-video | AI Lab | ↗ | |
| 33 | vidu-q2-turbo | Shengshu | ↗ | |
| 34 | kling-v2.1-master | KlingAI | ↗ | |
| 35 | hailuo-02-pro | MiniMax | ↗ | |
| 36 | kling-v2.1-standard | KlingAI | ↗ | |
| 37 | ray-3 | Luma AI | ↗ | |
| 38 | hailuo-02-standard | MiniMax | ↗ | |
| 39 | vidu-q2-pro | Shengshu | ↗ | |
| 40 | hunyuan-video-1.5 | Tencent | ↗ | |
| 41 | hailuo-02-fast | MiniMax | ↗ | |
| 42 | seedance-v1-lite | Bytedance | ↗ | |
| 43 | wan-v2.2-a14b | Alibaba | ↗ | |
| 44 | veo-2 | ↗ | ||
| 45 | ltx-2-19b | AI Lab | ↗ | |
| 46 | ray2 | Luma AI | ↗ | |
| 47 | runway-gen4-turbo | Runway | ↗ | |
| 48 | pika-v2.2 | Pika | ↗ |
视频编辑与二创风格榜
视频风格重绘与质感升级 (Video Edit) · 考核对已有视频进行风格迁移、主体替换、滤镜重构与高清质感升级的二次创作能力
视频编辑与二创风格榜 · 详细排名 (TOP 50)
arena.ai 实测排位| 排名 | 模型 | 机构 | 评分/战力 | 官网 |
|---|---|---|---|---|
| 🥇 1 | wan3.0 | Alibaba | ↗ | |
| 🥈 2 | dreamina-seedance-2.5-720p | Bytedance | ↗ | |
| 🥉 3 | minimax-h3 | MiniMax | ↗ | |
| 4 | gemini-omni-flash | ↗ | ||
| 5 | dreamina-seedance-2.0-720p | Bytedance | ↗ | |
| 6 | happyhorse-1.0 | Alibaba-ATH | ↗ | |
| 7 | grok-imagine-video | SpaceXAI | ↗ | |
| 8 | kling-o3-pro | KlingAI | ↗ | |
| 9 | kling-o1-pro | KlingAI | ↗ | |
| 10 | runway-gen4-aleph | Runway | ↗ |
更多榜单与全量明细速查
从下拉菜单选择任意分类,一键展开查阅该分项完整的实测排行榜数据