Pass IndexThe State of AISign in

Muse Spark 1.1

Muse Spark 1.1 is strongest at agentic performance, tool use, and computer use.

text + image + file → text · made by Meta

$1.25$4.25per Mtok in / out
Meta · 3 sellers

Sold by 4 ways

SellerLaneRate
Meta apistandard$1.25per Mtok in$4.25per Mtok out$0.15per Mtok cacheddeveloper.meta.com · read 2026-08-25
Nous Research aggregatorstandard$1.25per Mtok in$4.25per Mtok out$0.15per Mtok cachedinference-api.nousresearch.com · read 2026-08-25
OpenRouter 0 ways$1.25per Mtok in$4.25per Mtok out$0.15per Mtok cached
OpenRouter aggregatornot on the seller's list todaystandard$1.25per Mtok in$4.25per Mtok out$0.15per Mtok cachedopenrouter.ai · read 2026-08-24
Vercel AI Gateway aggregatorstandard$1.25per Mtok in$4.25per Mtok out$0.15per Mtok cachedai-gateway.vercel.sh · read 2026-08-25

Measured 40 standings

PlaceBoardMetricScore
1stof 25SWE-Bench Pro (Public Dataset)Resolve Rate61.5
6thof 28τ³-Banking LeaderboardPass^140.5
7thof 65Apex agents — Epoch AIPass@1 score0.419
7thof 110OSWorld-Verified Results (All)Success Rate (Avg±Std)80.7
7thof 168Scicode — Epoch AIScore58.218
8thof 17terminal-bench@2.1Accuracy (xhigh)76.2
10thof 92Design Arena · asciiart (models)Elo1301
11thof 402LMArena · TextScore (Elo)1493
11thof 60Vending bench 2 — Epoch AIScore6520.475
11thof 60Vending-Bench 2 — Current leaderboardMoney Balance (average across 5 runs)6520.47
12thof 36Design Arena · htmlslides (agents)Elo1197
13thof 23Gbaeval — Epoch AIOverall score0.079
14thof 80Simpleqa verified — Epoch AImean_score57.789
15thof 123Lmca — Epoch AIScore (high)49.91
16thof 26Surface evolver bench — Epoch AIMean score (high)0.525
17thof 161Dtbench — Epoch AIAccuracy (high)94.4
18thof 109Design Arena · svg (models)Elo1278
20thof 40Design Arena · agenticgamedev (agents)Elo1183
22ndof 152LMArena · VisionScore (Elo)1281
23rdof 39Gdp pdf — Epoch AIGDP.pdf score0.15
24thof 64Proofbench — Epoch AIAccuracy39
25thof 152Design Arena · dataviz (models)Elo1291
25thof 150Design Arena · uicomponent (models)Elo1308
26thof 155Design Arena · gamedev (models)Elo1300
27thof 121Webdev arena — Epoch AIArena Score1539.02
28thof 156Design Arena · codecategories (models)Elo1292
28thof 64Design Arena · fullstack (agents)Elo1212
28thof 63Design Arena · mobileapps (agents)Elo1203
29thof 128LMArena · WebDevScore (Elo)1542
30thof 147Design Arena · 3d (models)Elo1293
30thof 42Design Arena · godotgamedev (agents)Elo1141
30thof 62Design Arena · webapps (agents)Elo1217
31stof 55Design Arena · androidnative (agents)Elo1189
32ndof 43Design Arena · python-pptxslides (agents)Elo1156
35thof 162Design Arena · website (models)Elo1282
39thof 184Artificial Analysis · Coding IndexIndex71.3
45thof 69Deepswe — Epoch AIPass@10.533
52ndof 177Critpt — Epoch AIAccuracy15.1
57thof 143Artificial Analysis · Agentic IndexIndex27.5
143rdof 656Epoch capabilities index — Epoch AIECI Score154.65

About

Muse Spark 1.1 — a text model from Meta, sold by 3 companies from $1.25 in and $4.25 out per million tokens, placed 1st of 25 on SWE-Bench Pro (Public Dataset).

It takes text, images and files and returns text, with a context window of 1,048,576 tokens. It was published in April 2026. Its sellers say it can reason step by step and call a tool. The catalogue files it under code and agents. Three companies sell it. The cheapest is $1.25 in and $4.25 out per million tokens at Meta. It has been measured on 40 boards, and stands best at 1st of 25 on SWE-Bench Pro (Public Dataset).

Every current figure

Maker
Meta
Register
model
Takes
text + image + file
Returns
text
Context
1,048,576 tokens
Published
April 2026
Licence
not read
Sellers
3
Price
$1.25 in and $4.25 out per million tokens — Meta
Boards
40
Best place
1st of 25 — SWE-Bench Pro (Public Dataset)

Known as 4 names

muse-spark-1.1Muse Spark 1.1*mini-SWE-agent + Muse Spark 1.1 (xhigh)meta/muse-spark-1.1