OpenAI’s image generation model released April 2026 (also called ChatGPT Images 2.0). Ranked #1 on arena.ai as of April 2026, beating Imagen 3 (Nano Banana 2) by 24 points — the largest benchmark gap ever recorded at that evaluation.
Key facts
- Type: Image generation model
- Maker: OpenAI
- Released: April 2026
- Also known as: ChatGPT Images 2.0
- Pricing: $0.06/image flat
- Benchmark: #1 on arena.ai, +24 points over Imagen 3 (largest ever gap at this benchmark)
- Availability: ChatGPT, direct API, key.ai aggregator
Strengths
- Text rendering (best-in-class)
- Photorealism
- Product packaging
- Consistent character generation across multiple images
- Ad creative generation
Weaknesses vs Imagen 3
- Imagen 3 still wins on some product photography scenarios (uses web search to pull real logos for accuracy) [012]
- Roughly equivalent for diagrams and some product shots [012]
Related
- See also: Openai, Imagen 3, Key Ai
- Concepts: Model Benchmarking Methodology
Source references
- [012] Nate Herk — Video editing & content creation cluster (2026-04-15 to 2026-04-23)
Recommended next
Keep reading from this thread
From 491 indexed pages and articles.
- Wiki concept GPT 5.5 OpenAI's flagship model released April 2026 (codenamed "Spud"). Positioned as "more with less" — fewer output tokens per task, faster autonomous decomposition of vague Related by gpt
- Wiki concept OpenAI Whisper Speech-to-text model family used for video transcription and realtime speech recognition. In the wiki it first appears as the transcription backend for AI video Related by gpt
- Insight Recommendation Systems in Production How recommendation systems become production decisioning systems through signals, ranking, constraints, feedback loops, and experimentation Readers have engaged with this next