When was Muse Spark 1.2 released?

Meta released Muse Spark 1.2 on 5 Aug 2026. It is available through the qynio gateway.

Does Muse Spark 1.2 support zero data retention?

Not by default. See the per-route data-handling section for each route's logging stance and retention window.

Where is Muse Spark 1.2 hosted?

Muse Spark 1.2 is available on multi-region hosting through the qynio gateway. See the per-route table for details.

Is Muse Spark 1.2 trained on prompts and outputs?

No. Every hosting route for Muse Spark 1.2 is configured so customer prompts and outputs are not used to train models by default.

Is a no-KYC deposit available for Muse Spark 1.2?

Yes. A no-KYC Data Processing Agreement is available on at least one route serving Muse Spark 1.2. Contact hello@qyn.io to receive it.

Muse Spark 1.2

by Meta

Released 607 days after Llama 3.3 70B · All Meta releases

Muse Spark 1.2 arrived on August 5, 2026 as the coding-focused flagship of Meta Superintelligence Labs' Muse family, following the original Spark in April and the 1.1 refresh in July. The model keeps the line's million-token context window and multimodal input across text, images, audio, and video, while sharpening first-attempt coding accuracy and long multi-step reasoning. Debugging sprawling codebases, validating changes that touch thousands of files, and orchestrating parallel subagents are the workloads it was built for, and it also powers Muse Code, Meta's terminal coding agent. Tool use, MCP integrations, and persistent background agents make it a strong pick for deep agentic software engineering pipelines.

Open in playground Try in chat Compare

Key info

Input

Output

Features

Context window

Max output

—

Input price

$1.25 /1M

Output price

$4.25 /1M

Released

5 Aug 2026

No training by default
no-KYC deposit available

Available routes

Muse Spark 1.2 runs on 1 route through the qynio gateway. Compare residency, zero retention, and training posture at a glance — full data-handling detail per route below.

Provider	Region	Zero data retention	Training	Input	Output
Meta	Multi	Not by default	No	$1.25	$4.25

Data handling per route

Each route hosting Muse Spark 1.2 has its own privacy posture, residency, and no-KYC terms. Postures are maintained by qynio with a last-verification timestamp.

Meta — Multi-region

Zero data retention is not on by default — abuse monitoring. No training on customer data. GLOBAL; wallets; deposit available.

Zero data retention: Not by default — abuse monitoring. Derived from the logging and moderation facts.
Training: No training on customer data.
Logging: Abuse monitoring
Moderation: Not established
Caching: Not established
Subprocessor access: Not established
no-KYC deposit: deposit available
Transfer mechanism: wallets

Benchmarks

Independent benchmark scores — composite indices for reasoning, coding, and math, plus individual eval scores where available.

Global rank#30 of 630 LLMs

TierFrontier

Output speed224 tok/s

First token13.36s

Intelligence Index56.8

Coding Index72.2

Reasoning & knowledge

GPQA Diamond

90%

Humanity's Last Exam

46%

Long-context reasoning

83%

Coding

SciCode

56%

See full leaderboard →Benchmarks via Artificial Analysis · View on AA

Get started

Call Muse Spark 1.2 through the qynio gateway with one API key. Let your coding agent set it up, or call it directly — qynio is drop-in compatible with the OpenAI, Anthropic, and Google AI SDKs.

Set it up with your agent

Copy this and paste it into a coding agent like Claude Code, Cursor or Codex and it'll wire up qynio for you.

Or call it directly

1import OpenAI from "openai";
2
3const client = new OpenAI({
4  apiKey: process.env.QYN_API_KEY,
5  baseURL: "https://api.qyn.io/v3/compat",
6});
7
8const completion = await client.chat.completions.create({
9  model: "meta/muse-spark-1.2",
10  messages: [{ role: "user", content: "Hello" }],
11});
12console.log(completion.choices[0].message.content);