Muse Spark 1.3

+ contribute

Muse Spark 1.3 is a closed-weights language model from Meta, released September 2026.

77.2±0.8#24 of 719 LLMsfrom 14 benchmark results across 5 sources

$ how it compares

The LLMs ranked either side of Muse Spark 1.3, out of 828 with a score, then more from Meta.

#	model	score	where	RAM	released
22	Claude Haiku 5.5 Anthropic	77±1	api $0.20/M		2026-10
23	Gemini 3.8 Flash Google	77±1	api $1.50/M		2026-09
24	Muse Spark 1.3 Meta	77±1	api $2.00/M		2026-09
25	Kimi K3 Moonshot AI	77±1	api $6.00/M +local	~884GB	2026-07
26	GPT-5.2 Pro OpenAI	77±1	api $57.75/M		2025-12
	more from Meta
40	Muse Spark 1.2 Meta	75±1	api $2.00/M		2026-08
43	Muse Spark 1.1 Meta	75±1	api $2.00/M		2026-07
44	Muse Spark Meta	75±2	comparison only		2026-04

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▁ #29 of 300 · 20% of its score · 3 tests

Beats GPT-5.4 here, though they rank higher overall. Trails GLM-5.3-Flash, Grok 4.7, Qwen3.8 Max (0902) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Arena Agent (October 2026 aggregate IPS)	0.0400 IPS ✎	▁▁▁▄▂▃▇█▂▂▁▃ beats 33/49	avg -0.312 IPS	▇▇▇▇ 10%	sharp
Terminal-Bench 4.0	14.5% ✎	█▅▅▅▅·▅▃▃▅▅▅ beats 3/20	very hard · avg 0.4%	▇▇ 6.0%	sharp · at its level
E2E-SWE (build a working codebase from scratch, 186 tasks) pass@1 lab	45.2% ✎	▃··▅··▃█▃▅▃▅ beats 6/12	very hard · avg 3.1%	▇▇ 3.9%	sharp · at its level

coding▁▁▁▃▆██▇▆▄▃▁ #12 of 458 * · 13% of its score · 2 tests

Beats GPT-5.6 Sol, Claude Fable 5, GPT-5.5 here, though they rank higher overall. Trails Qwen3.8 Max (0902) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Arena WebDev (Code Arena)	1656.6 ✎	▁▂▄▃▅█▄▇▆▄▁▁ beats 108/119	avg 981	▇▇▇▇ 9.0%	sharp
E2E-SWE (build a working codebase from scratch, 186 tasks) pass@1 lab	45.2% ✎	▃··▅··▃█▃▅▃▅ beats 6/12	very hard · avg 3.1%	▇▇ 3.9%	sharp · at its level

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #8 of 393 * · 5.0% of its score · 1 test

Beats GPT-6 Astra, GPT-6.1 Sol, Claude Sonnet 5.5 here, though they rank higher overall. Trails Claude Opus 4.6, Claude Opus 4.7 here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
LMArena (text)	1494.3 ✎	▁▁▁▂▃▃▃▆▆▇█▃ beats 372/379	avg 1268	▇▇ 5.0%	blunt

math▁▂▃▅▆▅▄█▄▃▁▁ #22 of 398 · 48% of its score · 5 tests

Beats DeepSeek V4.1 Flash, Gemini 3.8 Flash here, though they rank higher overall. Trails GPT-5.6 Luna, GPT-6 Luna, GPT-5.2 Pro here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
FrontierMath tier 4 v2	46.3% ✎	▁▂▁▂█▅▂▂▁▁▁▁ beats 46/63	very hard · avg 0.1%	▇▇▇▇▇▇▇▇▇ 23%	sharp · at its level
FrontierMath tiers 1–3 v2	74.0% ✎	▂··▁▁▃▄▄█▅▃▃ beats 63/82	very hard · avg 1.2%	▇▇▇▇▇▇▇▇ 20%	sharp · at its level
MathArena ArXivMath August 2026, with tools (57 problems)	45.6% ✎	█▄·▄·······█ beats 2/5	very hard · avg 4.8%	▇ 2.6%	sharp · at its level
MathArena BrokenArXiv August 2026, with tools (56 problems)	20.2% ✎	▄▄·····▄·▄·█ beats 0/5	very hard · avg 5.5%	▇ 1.1%	at its level · noisy
OTIS mock AIME 2024–2025	99.2% ✎	▃▃▄▃▃▃▇▇█▃▂▅ beats 182/193	very hard · avg 8.5%	▇ 0.6%	sharp · maxed out · noisy

vision & documents▁▁▃▂▁▃▄▂▆▇█▄ #17 of 147 * · 5.2% of its score · 2 tests

Beats Claude Sonnet 5.5, GPT-5.5, MiMo-V2.6-Pro here, though they rank higher overall. Trails Gemini 3.5 Flash, Claude Opus 4.7, Muse Spark 1.2 here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Arena vision (style controlled)	1290.2 ✎	▁▁▂▂▂▂▄▅▅▆██ beats 136/145	avg 1068	▇ 3.5%	blunt
Arena document (style controlled)	1467.9 ✎	▁▁▃▃▄▃▇█▅▇▄▅ beats 23/37	avg 1335	▇ 1.7%	blunt

other tests2 not in a task · 13% of its score

test	result	vs the others in its run	difficulty	counts	why
Mystery Game Puzzles (Epoch)	25.0% ✎	█··▂▃▇▄█▂▃▁▁ beats 43/74	very hard · avg 9.4%	▇▇▇ 8.0%	sharp · at its level · noisy
Chess Puzzles (Epoch)	38.0% ✎	█▁▁▁▁▂▁▄▂▃▁▁ beats 115/135	very hard · avg 5.9%	▇▇ 4.7%	sharp · at its level · noisy

$ where to run it

Through an API

api · $2.00/M · $1.25 in / $4.25 out per 1M tokens · 1M context · dev.meta.ai ↗ · price sourceUSD per 1M tokens, Meta Developer Platform Standard tier (prompts not used for training) for muse-spark-1.3; cached input $0.15/1M and the cheaper Contributor tier (muse-spark-1.3-contributor, data used for training) excluded; web-search grounding billed separately; page states no long-context premium. Currency: USD. Meta official pricing page, observed 2026-09-25.

$ how the score was worked out · Epoch AI — internal evaluatio… 57%, LMArena leaderboard dataset 29%, Terminal-Bench 4.0 official l… 6% and 2 more

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

source	results	share of the score	licence
Epoch AI — internal evaluations	5	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 57%	CC BY 4.0
LMArena leaderboard dataset	5	▇▇▇▇▇▇▇▇▇ 29%	CC BY 4.0
Terminal-Bench 4.0 official leaderboard submissions	1	▇▇ 6%	Apache-2.0
E2E-SWE: Benchmarking LLMs on Building Working Codebases from Scratch (arXiv:2609.38335v1) · lab-reported	1	▇ 4%	CC BY 4.0
MathArena ArXivMath / BrokenArXiv / ArXivLean model-output datasets	2	▇ 4%	CC BY-SA 4.0
prior · pull toward the middle		▇ 0%
weighted average		= 77.2