Mind Lab: Macaron V1 Tall
mindlab/macaron-v1-tall
Try Macaron V1 Tall for free in your browser, and add it to your app for free with Puter.js AI API.
Try it free Add to your app// npm install @heyputer/puter.js
import { puter } from '@heyputer/puter.js';
puter.ai.chat("Explain quantum computing in simple terms", {
model: "mindlab/macaron-v1-tall"
}).then(response => {
document.body.innerHTML = response.message.content;
});
<html>
<body>
<script src="https://js.puter.com/v2/"></script>
<script>
puter.ai.chat("Explain quantum computing in simple terms", {
model: "mindlab/macaron-v1-tall"
}).then(response => {
document.body.innerHTML = response.message.content;
});
</script>
</body>
</html>
Model Card
Macaron V1 Tall is a text model from Mind Lab, the maker of the Macaron models. It uses a Mixture of LoRA (MoL) design on a Qwen3.6-35B-A3B base, with four specialists for chat, agents, coding and generative UI (UI4A). A router picks one specialist for each request. It supports tool calling and reasoning, with a 262,144-token context window and up to 65,536 output tokens.
In Mind Lab's reported results against the Qwen3.6-35B-A3B base, Macaron V1 Tall scores 75.4 on SWE-bench Verified (base 73.4), 56.2 on TerminalBench 2.1 (base 52.5) and 59.3 on UI4A-Bench (base 33.9). Mind Lab describes this as a system comparison, not a controlled ablation.
It suits chat assistants, agent workflows with tool use, repository-level coding and interface generation. It is the smaller sibling of Macaron V1 Venti, and Mind Lab reports an average routing loop of 1.76 s against 4.68 s for Venti.
Context Window 262K
tokens
Max Output 66K
tokens
Input Cost N/A
per million tokens
Output Cost N/A
per million tokens
Input text
modalities
Tool Use Yes
Release Date Oct 15, 2026
Try Macaron V1 Tall for free
Try Macaron V1 Tall instantly in your browser.
This playground uses the Puter.js AI API — no API keys or setup required.
More AI Models From Mind Lab
Macaron V1.1 Venti
Macaron V1.1 Venti is a text model from Mind Lab that combines a shared base model with four LoRA specialists for chat, agent work, coding and generative UI. It follows Macaron V1 Venti and moves the base from GLM-5.2 to GLM-5.3. It supports tool calling, a 1,048,576-token context window and up to 131,072 output tokens. Mind Lab reports ChatBench v2 at 66.7 (V1 at 63.6), Toolathlon-Verified at 76.0 (V1 at 63.0), DeepSWE v1.1 at 71.7 (V1 at 58.4) and UI4ABench v2 at 83.2 (V1 at 75.8). These are vendor-reported, and the model card says some competitor scores come from different protocols. It suits chat assistants, tool-using agents, software engineering tasks and interfaces generated from prompts through the API.
ChatMacaron V1.1 Coding Venti
Macaron V1.1 Coding Venti is a coding model from Mind Lab, the maker of the Macaron models. It is the coding variant of Macaron V1.1, which is post-trained from GLM-5.3 and combines a shared base with Chat, Agent, Coding, and Generative UI LoRA specialists. The coding specialist targets long-horizon software engineering and terminal tasks. The model accepts text and image input, returns text, supports tool calling, and has a 1,048,576-token context window with up to 131,072 output tokens. Mind Lab reports scores for the general V1.1 model, not this variant alone. These include 71.7 on DeepSWE v1.1 (Claude Opus 5 at 68.8), 50.0 on SWE-Marathon (Opus 5 at 50.0), and 31.4 on Terminal-Bench 3.0 (Opus 5 at 42.7). It suits repository-level bug fixing, terminal agents, and coding assistants.
Frequently Asked Questions
You can access Macaron V1 Tall by Mind Lab through Puter.js AI API. Include the library in your web app or Node.js project and start making calls with just a few lines of JavaScript — no backend and no configuration required. You can also use it with Python or cURL via Puter's OpenAI-compatible API.
Macaron V1 Tall is free to try with a Puter account. Every account includes a free AI allowance, and you can chat with it in the playground on this page. You can upgrade your account anytime for a larger allowance.
Macaron V1 Tall is free to integrate using the Puter.js AI API. With the User-Pays Model, you can add AI to your app for $0, since users cover their own AI usage through their Puter account.
Macaron V1 Tall was created by Mind Lab and released on Oct 15, 2026.
Macaron V1 Tall supports a context window of 262K tokens. For reference, that is roughly equivalent to 524 pages of text.
Macaron V1 Tall can generate up to 66K tokens in a single response.
Macaron V1 Tall accepts the following input types: text. It produces: text.
Yes, Macaron V1 Tall supports tool use (function calling), allowing it to interact with external tools, APIs, and data sources as part of its response flow.
Yes — the Macaron V1 Tall API works with any JavaScript framework, Node.js, or plain HTML through Puter.js. Just include the library and start building. See the documentation for more details.
Add Macaron V1 Tall to your app for free
Developers can integrate Macaron V1 Tall for free using the Puter.js AI API.
With the User-Pays Model, each user covers their own AI usage instead of the developer.