Models by task
Choose a model and Mac memory budget for your task. This guide combines official information and user feedback; measurements have their own page.
Site development and operationsTest feature changes, bug fixes and site checks using a repository and its tests.Qwen3-Coder-30B-A3B-Instruct · 32 GBModels and capabilities Voice inputTurn speech into text ready to enter into an app. Compare with Typeless and Aqua Voice and check where each processes audio.Whisper large-v3-turbo · 16 GBModels and capabilities Documents, search and meeting notesSummarize PDFs and documents, find evidence in local files, and compare workflows for transcripts and meeting notes.Qwen3.5-9B · 16 GBModels and capabilities Decisions and classificationCheck publication criteria, route incoming messages, and decide whether a document is relevant to a question.Qwen3.5-9B · 16 GBModels and capabilities Image and comic generationBuild on the existing image tests by checking the complete workflow for four-panel and explanatory comics.FLUX.2 [klein] 4B · 24 GBModels and capabilities Writing and translationSee model answers, knowledge and arithmetic scores, response time, memory and English–Japanese translation examples and scores.Qwen3.5-9B · 16 GBModels and capabilities Video generationCompare clips made from the same prompts, their generation time and their memory use.Wan2.2 TI2V-5B · 64 GBModels and capabilities
Starting memory estimate
Memory figures are editorial budgets including room for the OS and apps, not verified minimums. Text models assume 4-bit weights, 4K–8K input tokens and one model at a time. Long documents, images and concurrent apps need more. A3B does not mean the entire model fits in 3GB.
Assumes an Apple Silicon Mac and a compatible app: LM Studio or similar for text, Whisper apps for audio, and model-specific image/video tools. RAM alone does not determine speed. Start with a short input on your current Mac.
| Starting memory estimate | Model | Format |
|---|---|---|
| 16 GB | Qwen3.5-9B | 4bit / Q4_K_M |
| 32 GB | Qwen3-Coder-30B-A3B-Instruct | 4bit / Q4_K_M |
| 48 GB | Qwen3.6-35B-A3B | 4bit / Q4_K_M |
| 16 GB | Whisper large-v3-turbo | MLX / FP16 |
| 24 GB | FLUX.2 [klein] 4B | distilled / BF16 or app-supported quantization |
| 64 GB | Wan2.2 TI2V-5B | app-specific / offload |