What is the cost of standard input tokens for Muse Ember 1.0?

Standard input tokens under 200K cost $0.10 per 1M tokens, while cached input tokens cost $0.004 per 1M tokens.

What is the maximum context window for Muse Ember 1.0?

The model handles a context window of approximately 1,048,576 tokens and outputs up to 524,288 tokens.

What modalities does Muse Ember 1.0 support?

Inputs accept text, image, video, audio, and PDF data, while outputs are strictly text.

Muse Ember 1.0 Pricing and Specs Guide

Muse Ember 1.0 API Pricing and Benchmarks. A macro shot of a precision-machined copper cold plate resting on a dark server chassis, reflecting the amber glow of a single status LED on a distant rack — GPUniq

Data as of November 10, 20263 sources

Drafted with AI tools and edited by a human. The figures and conclusions were checked by a GPUniq editor.

TL;DR

Muse Ember 1.0 is Meta's smallest efficient reasoning model. It costs $0.10 per 1M standard input tokens and $0.50 per 1M output tokens. Cached input tokens drop to $0.004 per 1M. The model handles a 1M token context window and outputs up to 524,288 tokens. It scores 36 on the Artificial Analysis Index v5.0.

API pricing

Standard input tokens under 200K cost $0.10 per 1M. Output tokens are priced at $0.50 per 1M. Cached input tokens cost $0.004 per 1M. These rates apply on the Meta Model API in standard mode.

Item	USD Rate	Unit
Input tokens (standard mode, <=200K input tokens)	0.10	per 1M tokens
Output tokens	0.50	per 1M tokens
Cached input tokens	0.004	per 1M tokens
Long-context input tokens (>200K input tokens)	0.30	per 1M tokens
Long-context output tokens (>200K input tokens)	1.50	per 1M tokens

Long context rates

Inputs exceeding 200K tokens cost $0.30 per 1M. Outputs past that threshold cost $1.50 per 1M.

Context limits

The context window spans approximately 1,048,576 tokens. Maximum output reaches 524,288 tokens.

Benchmark scores

The model scores 36 on the Artificial Analysis Index v5.0. OSWorld 2.1 hits 69.5 percent. PerceptionBench with tools reaches 80.6 percent.

Benchmark	Score
Artificial Analysis Index v5.0	36
OSWorld 2.1	69.5%
PerceptionBench (with tools)	80.6%

Modality support

Inputs accept text, image, video, audio, and PDF data. Outputs are strictly text.

Methodology

Facts are compiled from the vendor's own announcement and documentation, linked below, at the time of writing; where the vendor hasn't published a number it is marked as such. GPUniq prices, when the model is already available here, come from our live catalog.