AI Leaderboard

41 models benchmarked · updated April 9, 2026

Agentic Coding

🥇GPT-5.4 (xhigh)OpenAI57.3
🥈Gemini 3.1 Pro PreviewGoogle55.5
🥉Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)Anthropic50.9
4Muse SparkMeta47.5
5Qwen3.6 PlusAlibaba42.9
6Grok 4.20 0309 (Reasoning)xAI42.2
7DeepSeek V3.2 SpecialeDeepSeek37.9
8NVIDIA Nemotron 3 Super 120B A12B (Reasoning)NVIDIA31.2
9Mistral Small 4 (Reasoning)Mistral24.3

Conversation

🥇Gemini 3.1 Pro PreviewGoogle57.2
🥈GPT-5.4 (xhigh)OpenAI57.2
🥉Claude Opus 4.6 (Adaptive Reasoning, Max Effort)Anthropic53.0
4Muse SparkMeta52.1
5Qwen3.6 PlusAlibaba50.0
6Grok 4.20 0309 v2 (Reasoning)xAI49.3
7DeepSeek V3.2 (Reasoning)DeepSeek41.7
8NVIDIA Nemotron 3 Super 120B A12B (Reasoning)NVIDIA36.0
9Mistral Small 4 (Reasoning)Mistral27.2

Image Editing

🥇GPT Image 1.5 (high)OpenAI1271
🥈Nano Banana Pro (Gemini 3 Pro Image)Google1251
🥉grok-imagine-imagexAI1225
4Wan 2.6 ImageAlibaba1195

Image to Video

🥇grok-imagine-videoxAI1331
🥈Veo 3.1 FastGoogle1289
🥉Wan 2.5 PreviewAlibaba1252
4SoraOpenAI977

Text to Image

🥇GPT Image 1.5 (high)OpenAI1265
🥈Nano Banana 2 (Gemini 3.1 Flash Image Preview)Google1257
🥉grok-imagine-imagexAI1171
4Qwen Image Max 2512Alibaba1149
5Sana Sprint 1.6BNVIDIA935
6Janus ProDeepSeek704

Text to Speech

🥇TTS-1OpenAI1102
🥈Gemini 2.5 Flash Lite TTSGoogle1080
🥉Voxtral TTSMistral1029
4Magpie-Multilingual 357MNVIDIA1001
5Qwen3 TTSAlibaba932

Text to Video

🥇grok-imagine-videoxAI1229
🥈Veo 3.1 PreviewGoogle1227
🥉Sora 2 ProOpenAI1195
4Wan 2.6Alibaba1187

Spoton orchestrates all of this for you

Why pick one when you can have the best of every AI, for every task?