Models
Overviews of the key AI models: text, images, video and music.
ClaudeLong context, careful answers and strong work with code.SunoA song with vocals and arrangement in a single generation.Nano BananaEdit an image in plain words — and the character stays the same in every frame.ChatGPTThe GPT family: from fast nano tiers to the largest models and Codex for code.GPT ImageFollows a long request to the letter and sets legible text on a layout.KlingVideo from text and from a photo: smooth motion and sound in frame.GeminiA million tokens on every tier and best-in-class PDF reading.GrokUp to 2M tokens of context, fresh data from the web and image generation.VeoVideo with sound in a single render — speech, ambience and music already in the frame.SeedanceSeveral shots in one generation — the scene arrives whole.CohereAnswers from your own document base and shows which file each part came from.DeepSeekFlagship-level quality at a fraction of the price. Open weights.GigaChatNative-level Russian and a 128K context in the top tiers.GLMModels tuned for code and agentic work at well below market price.KimiHolds a million tokens in context, and almost the whole family is open.LLaMAMeta's open models: a free licence and a low price.MiMoA million-token context and reasoning at one of the lowest prices around.MiniMaxLong context, reasoning and open weights well below market price.MistralEuropean engineering, compact models and image understanding.PerplexityAnswers from a fresh search, and shows where every fact came from.QwenDozens of models from compact to flagship, and almost all of them open.YandexGPTNative-level Russian and a firm grip on regional context.FLUXImages and video from one family: photorealism, text in frame and sound in the clip.MidjourneyThe most recognisable look in generative imagery — no subscription, no Discord.
