Head-to-Head Leaderboard
A persistent cross-model leaderboard ranking AI models by Elo and win rate across every head-to-head matchup, with separate ladders for text, image, and video.
Text models
| Rank | Model | Elo | Win rate | Matches |
|---|---|---|---|---|
| 1 | Anthropic: Claude Sonnet 5.5 | 1545 | 100% | 3 |
| 2 | grok-4.6 | 1545 | 100% | 3 |
| 3 | Anthropic: Claude Fable 5.1 | 1544 | 80% | 5 |
| 4 | Anthropic: Claude Opus 5.5 | 1532 | 100% | 2 |
| 5 | MoonshotAI: Kimi K3 | 1532 | 100% | 2 |
| 6 | Meta: Muse Spark 1.3 | 1531 | 100% | 2 |
| 7 | gpt-oss-120b | 1526 | 61% | 9 |
| 8 | DeepSeek-V4.1-Flash | 1517 | 75% | 2 |
| 9 | Qwen: Qwen3.8 Max (0902) | 1517 | 100% | 1 |
| 10 | Qwen: Qwen3.8 Omni Flash | 1516 | 100% | 1 |
| 11 | grok-4.7 | 1515 | 75% | 2 |
| 12 | muse-spark-1.3 | 1513 | 63% | 4 |
| 13 | Cohere-command-a-plus-05-2026 | 1509 | 60% | 5 |
| 14 | Qwen: Qwen3.8 Max Prime | 1501 | 50% | 2 |
| 15 | Kimi-K3 | 1501 | 50% | 1 |
| 16 | Qwen: Qwen3.8 2.4T A95B | 1500 | 50% | 3 |
| 17 | Google: Gemini 4 Argon | 1498 | 50% | 2 |
| 18 | Mistral-Large-3 | 1498 | 50% | 2 |
| 19 | Gemini 3.8 Flash | 1485 | 25% | 2 |
| 20 | Z.ai: GLM 5.3 Prime | 1485 | 33% | 3 |
| 21 | SpaceXAI: Grok 4.7 | 1484 | 0% | 1 |
| 22 | Z.ai: GLM 5.3 FlashX | 1484 | 0% | 1 |
| 23 | DeepSeek: DeepSeek V4 Pro 0813 | 1484 | 25% | 2 |
| 24 | MiniMax: MiniMax M2.7 | 1483 | 0% | 1 |
| 25 | cohere-command-a | 1475 | 30% | 5 |
| 26 | Phi-4-multimodal-instruct | 1474 | 17% | 3 |
| 27 | Mistral: Mistral Large 4 | 1470 | 0% | 2 |
| 28 | Codestral-2501 | 1464 | 25% | 6 |
| 29 | Llama-4-Scout-17B-16E-Instruct | 1461 | 25% | 6 |
| 30 | MiniMax: MiniMax M3 | 1455 | 0% | 3 |
| 31 | DeepSeek-V4-Pro | 1445 | 34% | 16 |
| 32 | Phi-4 | 1435 | 8% | 6 |
| 33 | Phi-4-mini-instruct | 1430 | 0% | 5 |
Image models
| Rank | Model | Elo | Win rate | Matches |
|---|---|---|---|---|
| 1 | Meta Muse Image Edit | 1584 | 100% | 6 |
| 2 | Fibo Bbq Preview | 1580 | 100% | 7 |
| 3 | Google: Nano Banana 2 (Gemini 3.1 Flash Image) | 1562 | 83% | 6 |
| 4 | Flux 3 Image | 1561 | 83% | 6 |
| 5 | Fibo | 1558 | 92% | 6 |
| 6 | Imagineart 2.0 Preview | 1553 | 92% | 6 |
| 7 | Fibo Lite | 1548 | 79% | 7 |
| 8 | Qwen Image 3 Image Editing | 1546 | 100% | 3 |
| 9 | OpenAI: GPT-5.4 Image 2 | 1541 | 75% | 6 |
| 10 | Ideogram V4.0 Text to Image | 1540 | 83% | 6 |
| 11 | Seedream 5.0 Pro Image Editing | 1533 | 63% | 12 |
| 12 | Cosmos 3 Super | 1518 | 71% | 7 |
| 13 | Krea 2 Text to Image Turbo LoRA | 1517 | 100% | 1 |
| 14 | Google: Nano Banana Pro (Gemini 3 Pro Image) | 1516 | 60% | 5 |
| 15 | GPT Image 2.5 Sunburst Edit | 1516 | 100% | 1 |
| 16 | Fibo Gen 1.5 Text to Image | 1514 | 67% | 3 |
| 17 | Krea 2 Large | 1509 | 63% | 8 |
| 18 | Bitdance | 1503 | 50% | 2 |
| 19 | Ideogram Transparent | 1502 | 50% | 1 |
| 20 | Hidream I1 Dev | 1487 | 25% | 2 |
| 21 | Longcat Image | 1486 | 38% | 4 |
| 22 | Boogu Image | 1486 | 25% | 2 |
| 23 | Dreamshaper | 1486 | 0% | 1 |
| 24 | Hidream O1 Image | 1485 | 0% | 1 |
| 25 | Lumina Image 2 | 1485 | 0% | 1 |
| 26 | CogView | 1485 | 25% | 2 |
| 27 | Luma Photon Flash Reframe | 1484 | 0% | 1 |
| 28 | Ideogram V2 | 1484 | 0% | 1 |
| 29 | Kling Image | 1473 | 25% | 4 |
| 30 | DeepSeek Janus-Pro | 1473 | 0% | 2 |
| 31 | Ideogram V4.5 Edit | 1459 | 36% | 7 |
| 32 | Kling Image | 1458 | 20% | 5 |
| 33 | Wan | 1443 | 0% | 4 |
| 34 | Hidream I1 Full | 1440 | 0% | 4 |
| 35 | Bagel | 1310 | 8% | 24 |
| 36 | AuraFlow | 1272 | 9% | 32 |
Video models
| Rank | Model | Elo | Win rate | Matches |
|---|---|---|---|---|
| 1 | Happy Horse Video Edit | 1637 | 96% | 13 |
| 2 | Seedance 2 Image to Video | 1583 | 86% | 11 |
| 3 | Gemini Omni Flash | 1581 | 100% | 7 |
| 4 | Happy Horse 1.1 Image to Video | 1581 | 89% | 9 |
| 5 | MiniMax H3 Reference to Video | 1559 | 100% | 4 |
| 6 | FLUX 3 Image to Video | 1553 | 90% | 5 |
| 7 | Luma Ray 3.2 Image to Video | 1547 | 73% | 11 |
| 8 | LTX 2.5 Image to Video Fast | 1531 | 70% | 5 |
| 9 | Gemini Omni Flash 1.1 Image to Video | 1530 | 83% | 3 |
| 10 | Kling O3 Image to Video [Pro] | 1517 | 75% | 2 |
| 11 | Kling Video V3 Turbo Pro Image to Video | 1517 | 75% | 2 |
| 12 | Wan 3.0 Prime | 1516 | 100% | 1 |
| 13 | Seedance 2.5 US Reference to Video | 1516 | 100% | 1 |
| 14 | Wan v2.6 Image to Video | 1511 | 63% | 8 |
| 15 | Marey Realism V1.5 | 1509 | 60% | 10 |
| 16 | LTX 2.3 Video Fast | 1508 | 50% | 2 |
| 17 | PixVerse C1 Image To Video | 1506 | 50% | 1 |
| 18 | LongCat Video Distilled | 1504 | 50% | 1 |
| 19 | Bytedance Seedance V1 Pro Fast Image To Video | 1504 | 50% | 1 |
| 20 | Pika Image to Video (v2.2) | 1502 | 50% | 1 |
| 21 | PixVerse Lipsync | 1501 | 50% | 1 |
| 22 | Kandinsky5 | 1501 | 50% | 2 |
| 23 | Krea Wan 14b- Text to Video | 1500 | 50% | 2 |
| 24 | H3 Max VHS | 1500 | 50% | 2 |
| 25 | Kling Video | 1500 | 50% | 2 |
| 26 | Boreal | 1499 | 50% | 2 |
| 27 | Bytedance Seedance V1.5 Pro Image To Video | 1493 | 44% | 8 |
| 28 | Bernini-R Reference Edit Video | 1490 | 38% | 13 |
| 29 | Kling Video v2.6 Image to Video | 1490 | 25% | 2 |
| 30 | LTX Video-0.9.7 13B Distilled | 1489 | 0% | 1 |
| 31 | Ltx 2.3 Quality | 1486 | 0% | 1 |
| 32 | PixVerse V4 Image To Video | 1485 | 0% | 1 |
| 33 | LTX-2.3 22B Distilled | 1485 | 0% | 1 |
| 34 | Ovi | 1485 | 0% | 1 |
| 35 | H3 Max Retro Toon 70s | 1484 | 0% | 1 |
| 36 | LTX-2.3 22B | 1484 | 0% | 1 |
| 37 | Luma Ray 2 (Image to Video) | 1484 | 0% | 1 |
| 38 | MAGI-1 (Distilled) | 1484 | 0% | 1 |
| 39 | MiniMax (Hailuo AI) Video 01 | 1484 | 0% | 1 |
| 40 | Heygen | 1476 | 0% | 2 |
| 41 | Kandinsky5 Pro | 1471 | 0% | 2 |
| 42 | Heygen v5 Digital Twin | 1470 | 0% | 2 |
| 43 | Infinity Star | 1470 | 0% | 2 |
| 44 | CogVideoX-5B | 1410 | 0% | 7 |
| 45 | AnimateDiff | 1343 | 0% | 14 |
See every matchup on the Head-to-Head index.