Junie Local is free. The 64 GB M5 Mac is the price.

Qwen3.6-27B at four bits is roughly 13.5 GB of weights ^[1]. JetBrains asks for 64 GB of unified memory ^[4], close to five times that footprint ^[2], and the download is about 20 GB ^[3], leaving some 6.5 GB the weights alone do not account for ^[3]. Both gaps point the same direction, and the company's own engineering write-up says where: prefill.

With a cloud model, when the agent needs a file again it simply reads it again, because prefill is fast and the context window can be pruned to whatever is relevant [c13a]. On local inference it is not fast [c13a]. So JetBrains changed the loop to append every new request directly to the rolling context, keeping files the model has already read resident so the KV cache survives from one task to the next ^[13], and reordered the session-start payload so the engine can cache the whole prefix up to the user's request and reuse it across tasks in the same project ^[14]. Read that against the memory floor and it looks less like a model-size requirement than a working-set one. The 64 GB is not there to hold the weights. It is there to hold everything the agent has decided never to read twice.

The chip floor has the same origin. JetBrains says most of an agent's wall-clock time goes on prefill rather than generation ^[11], and that Junie Local starts at M5 because the Neural Accelerator carries 8-bit arithmetic instructions the M4 lacks, worth around 40% more prefill throughput ^[12]. That part is software, and JetBrains says the patch is going upstream as a PR to MLX-VLM ^[12]. Its engine sits on mlx-vlm and Apple's MLX ^[17]; Ollama moved its Apple Silicon engine onto MLX back in March ^[18]. If the patch lands, the do-it-yourself crowd inherits the same throughput without buying anything.

What the free tier actually buys is a harness fitted to one quantized model. Qwen 3.6 ignores Junie's request for XML-style progress blocks, so JetBrains scrapes the model's plain-text narration of its tool calls instead, and notes that this kind of adaptation is model-specific ^[15]. Optional LLM calls and multi-agent mode are switched off, the latter because sequential processing is what an M5 handles best ^[16]. Point Junie at your own model in Ollama, which has been possible for a while ^[28], and you are back to picking weights and tuning settings yourself.

The quality claim needs reading carefully too. Parity with Sonnet 4.5 at a 10,000-token reasoning limit ^[8] is the local model with reasoning disabled entirely, measured against cloud models that had it on ^[9], because reasoning cost two to three times the tokens for very little quality ^[9]. Qwen3.8 needed it on and ran roughly four times slower for the privilege ^[6]. Everything here is a speed budget wearing different clothes, and JetBrains is blunt about the bill: 64 GB on an M5 is what a 27B model costs to run well today, and it is the number they are working hardest to bring down, according to head of marketing Dmitry Savelev ^[24][25].