Muse Spark 1.3

API ID muse-spark-1.3

Muse Spark 1.3 is a Meta model, released September 2026. Standard price: $1.25 input and $4.25 output per 1M tokens.

Status: Active
Released: September 2026
Shutdown: none announced
Replacement: —

Meta doesn't publish shutdown dates. AIWire checks its model list daily and will flag muse-spark-1.3 if it disappears.

Benchmarks

Scores as each source published them

Comparing with Muse Spark 1.2, the previous model in this line. A score is compared only with the same benchmark from the same source, at the same setting where both have it. Full comparison: prices, limits, features

Arena

Arena	Muse Spark 1.3	Muse Spark 1.2	Change	Rank	Date	Source
Text By category Math 1503 rank 13 vs 1483 (+20, within range) at max effort (vs xhigh effort) Creative writing 1461 rank 28 vs 1449 (+12, within range) at max effort (vs xhigh effort) Instruction following 1484 rank 18 vs 1472 (+12, within range) at max effort (vs xhigh effort) Hard prompts 1516 rank 12 vs 1512 (+4, within range) at max effort (vs xhigh effort) Long prompts 1497 rank 20 vs 1494 (+3, within range) at max effort (vs xhigh effort) Multi-turn 1490 rank 22 vs 1510 (−20, within range) at max effort (vs xhigh effort) Expert 1529 rank 14 vs 1492 (+37) at max effort (vs xhigh effort)	1494 ±6	1495 ±7	−1 within range	11 of 415 16,731 votes	Oct 13, 2026	source
Text: Coding	1537 ±9	1538 ±13	−1 within range	14 of 410 4,357 votes	Oct 13, 2026	source
WebDev Other settings 1622 ±8 vs 1533 ±14 xhigh effort	1659 ±8	1533 ±14	+126	12 of 142 8,683 votes	Oct 13, 2026	source
Vision Other settings 1286 ±10 vs 1293 ±14 xhigh effort	1290 ±10	1293 ±14	−3 within range	12 of 165 4,550 votes	Oct 9, 2026	source
Document	1468 ±18	not rated		19 of 44 1,006 votes	Sep 13, 2026	source
Agent	0.036 ±0.008	-0.065 ±0.008	+0.101	16 of 54 73,452 sessions	Oct 14, 2026	source

Rated as muse-spark-1.3-max (max effort); Muse Spark 1.2 as Muse Spark 1.2 (xHigh) (xhigh effort). Some were rated at different settings; ratings on one arena share one scale, so the difference still reads directly. Text, WebDev, Vision and Document ratings come from people voting between two anonymous answers; Text, Vision and Document are style-controlled, as on arena.ai. Agent is Arena's separate score for real agent sessions, on its own scale. Data: Arena (arena.ai), CC BY 4.0.

Epoch AI

Benchmark	Muse Spark 1.3	Muse Spark 1.2	Setting	Run	Source
FrontierMath Tier 4 Other settings 41.5% vs not run xhigh effort	46.3%	not run	max effort	Sep 18, 2026	source
FrontierMath Tiers 1-3 Other settings 74.0% vs not run max effort	74.4%	not run	xhigh effort	Sep 16, 2026	source
OTIS Mock AIME 2024-2025	99.2%	not run	xhigh effort	Sep 17, 2026	source

Run independently by Epoch AI. Data: Epoch AI, Capabilities & benchmarking (epoch.ai/benchmarks), CC BY.

Price history

USD per 1M tokens, standard tier

We started recording this model's prices on Oct 5, 2026. The history grows with each daily check.

Show as a table

From	Type	Price
Oct 5, 2026	Input	$1.25
Oct 5, 2026	Output	$4.25

Prices

USD per 1M tokens, every type and band, by tier

Type	Price $/1M	In effect
Standard
input	$1.25
cached input	$0.15
output	$4.25

Changelog

Changes we detected on Meta's pages

No changes recorded yet.

Email alerts

Get an email if muse-spark-1.3 changes

We'll email you the same morning we spot a deprecation notice, a new or moved shutdown date, a price change, or the model leaving Meta's model list.

No password: one email to confirm, and you can unsubscribe any time. Privacy

Last checked Oct 15, 2026, 2:00 am ET · Sources: Meta models, Meta pricing