KI-Rangliste

41 Modelle benchmarked · aktualisiert am 26. April 2026

Agentic Coding

🥇GPT-5.5 (xhigh)OpenAI59.1
🥈Gemini 3.1 Pro PreviewGoogle55.5
🥉Claude Opus 4.7 (Non-reasoning, High Effort)Anthropic53.1
4DeepSeek V4 Pro (Reasoning, Max Effort)DeepSeek47.5
5Muse SparkMeta47.5
6Qwen3.6 Max PreviewAlibaba44.9
7Grok 4.20 0309 (Reasoning)xAI42.2
8NVIDIA Nemotron 3 Super 120B A12B (Reasoning)NVIDIA31.2
9Mistral Small 4 (Reasoning)Mistral24.3

Konversation

🥇GPT-5.5 (xhigh)OpenAI60.2
🥈Claude Opus 4.7 (Adaptive Reasoning, Max Effort)Anthropic57.3
🥉Gemini 3.1 Pro PreviewGoogle57.2
4Muse SparkMeta52.1
5Qwen3.6 Max PreviewAlibaba51.8
6DeepSeek V4 Pro (Reasoning, Max Effort)DeepSeek51.5
7Grok 4.20 0309 v2 (Reasoning)xAI49.3
8NVIDIA Nemotron 3 Super 120B A12B (Reasoning)NVIDIA36.0
9Mistral Small 4 (Reasoning)Mistral27.8

Bildbearbeitung

🥇GPT Image 1.5 (high)OpenAI1261
🥈Nano Banana Pro (Gemini 3 Pro Image)Google1243
🥉grok-imagine-imagexAI1215
4Wan 2.6 ImageAlibaba1183

Bild zu Video

🥇grok-imagine-videoxAI1327
🥈Veo 3.1 Fast PreviewGoogle1277
🥉Wan 2.5 PreviewAlibaba1242
4SoraOpenAI978

Text zu Bild

🥇GPT Image 2 (high)OpenAI1332
🥈Nano Banana 2 (Gemini 3.1 Flash Image Preview)Google1263
🥉grok-imagine-imagexAI1185
4Qwen Image Max 2512Alibaba1162
5Sana Sprint 1.6BNVIDIA942
6Janus ProDeepSeek713

Sprachsynthese

🥇Gemini 3.1 Flash TTSGoogle1204
🥈TTS-1OpenAI1101
🥉Magpie-Multilingual 357M (Feb 2026)NVIDIA1071
4Voxtral TTSMistral1049
5Qwen3 TTSAlibaba936

Text zu Video

🥇grok-imagine-videoxAI1230
🥈Veo 3.1 PreviewGoogle1220
🥉Wan 2.6Alibaba1186
4Sora 2 ProOpenAI1183

Spoton orchestriert das alles für Sie

Warum sich entscheiden, wenn Sie das Beste jeder KI für jede Aufgabe haben können?