Skip to main content

MapikaSupersonic LabsDecision matchup

Decider 4Bvs.Julia 1

Which model makes better decisions?

Decider 4B does better in independent tests, while Julia 1 is much smaller.

Independent tests
Decider 4Bahead2 of 2
Julia 10 of 2
Jev Decision Index, DecideBench
Popularity
Decider 4B43
Julia 139
out of 100
Weights and access
Decider 4Bopen, local
Julia 1open, local
Size
Decider 4B4 B
Julia 1smaller0.14 B
Parameters

Pick Decider 4B for

Jev Decision IndexDecideBench

Pick Julia 1 for

smaller

Shared tests

Independent evaluators and leaderboards only, the newest shared snapshot per evaluator. With 95 % ranges a lead counts when they do not overlap; for hit rates, when it holds in a 95 % test. Without any uncertainty figure at most “slightly ahead”.

Jev Decision Index

Decider 4B ahead

Decider 4B is clearly ahead in 22 of 37 subtests and slightly in 12, Julia 1 in none. 3 are level.

Home appliances · Decision Index 0.4.1

Decider 4B17.1
Julia 10.0

88 tasks, as of October 28, 2026

POP909 · Decision Index 0.4.1

Decider 4B10.0
Julia 10.2

2,000 tasks, as of October 28, 2026

SATA-Bench · Decision Index 0.4.1

Decider 4B31.6
Julia 11.0

1,650 tasks, as of October 28, 2026

Show all 37 subtests
SubtestDecider 4BJulia 1Verdict
Home appliances · Decision Index 0.4.117.10.0Decider 4B ahead
POP909 · Decision Index 0.4.110.00.2Decider 4B ahead
SATA-Bench · Decision Index 0.4.131.61.0Decider 4B ahead
MMLU-Pro · Decision Index 0.4.151.714.7Decider 4B ahead
API-Bank · Decision Index 0.4.160.418.3Decider 4B ahead
When2Call · Decision Index 0.4.175.923.6Decider 4B ahead
HellaSwag · Decision Index 0.4.194.029.9Decider 4B ahead
New Yorker · Decision Index 0.4.162.923.5Decider 4B ahead
GSM8K · Decision Index 0.4.140.616.9Decider 4B ahead
BFCL · Decision Index 0.4.188.037.5Decider 4B ahead
BBH · Decision Index 0.4.160.933.8Decider 4B ahead
BPoMP · Decision Index 0.4.183.448.8Decider 4B ahead
GPQA Diamond · Decision Index 0.4.141.325.5Decider 4B ahead
WinoGrande · Decision Index 0.4.180.750.7Decider 4B ahead
ChessBench · Decision Index 0.4.110.57.5Decider 4B ahead
MuSR · Decision Index 0.4.153.739.0Decider 4B ahead
CLadder · Decision Index 0.4.163.350.2Decider 4B ahead
Humicroedit · Decision Index 0.4.159.748.3Decider 4B ahead
cfcolor · Decision Index 0.4.160.550.4Decider 4B ahead
PhishNChips · Decision Index 0.4.157.849.9Decider 4B ahead
Habermas · Decision Index 0.4.140.635.4Decider 4B ahead
HoVer · Decision Index 0.4.156.350.3Decider 4B ahead
CLINC150 · Decision Index 0.4.186.612.3Decider 4B slightly ahead
iSarcasmEval · Decision Index 0.4.142.16.5Decider 4B slightly ahead
ACOS · Decision Index 0.4.119.23.1Decider 4B slightly ahead
BRIGHT · Decision Index 0.4.145.59.1Decider 4B slightly ahead
ToolRet · Decision Index 0.4.165.013.6Decider 4B slightly ahead
FinEntity · Decision Index 0.4.186.032.8Decider 4B slightly ahead
ContractNLI · Decision Index 0.4.172.331.8Decider 4B slightly ahead
Amazon ESCI · Decision Index 0.4.142.420.6Decider 4B slightly ahead
ANLI · Decision Index 0.4.156.428.6Decider 4B slightly ahead
BANKING77 · Decision Index 0.4.174.742.6Decider 4B slightly ahead
VAST · Decision Index 0.4.153.532.9Decider 4B slightly ahead
NLI4CT · Decision Index 0.4.173.749.7Decider 4B slightly ahead
HLE · Decision Index 0.4.110.414.4level
CRUXEval · Decision Index 0.4.143.738.1level
RAGTruth · Decision Index 0.4.152.252.0level

Source · latest October 28, 2026

DecideBench

Decider 4B ahead

Decider 4B is clearly ahead in 10 of 10 subtests, Julia 1 in none.

DecideBench pair accuracy

Decider 4B78.5 %
Julia 18.0 %

200 tasks, as of October 18, 2026

DecideBench review_sentiment accuracy

Decider 4B88.0 %
Julia 122.0 %

50 tasks, as of October 18, 2026

DecideBench claim_support accuracy

Decider 4B90.0 %
Julia 124.0 %

50 tasks, as of October 18, 2026

Show all 10 subtests
SubtestDecider 4BJulia 1Verdict
DecideBench pair accuracy78.5 %8.0 %Decider 4B ahead
DecideBench review_sentiment accuracy88.0 %22.0 %Decider 4B ahead
DecideBench claim_support accuracy90.0 %24.0 %Decider 4B ahead
DecideBench returns_policy accuracy82.0 %28.0 %Decider 4B ahead
DecideBench overall accuracy88.5 %34.8 %Decider 4B ahead
DecideBench ticket_triage accuracy86.0 %34.0 %Decider 4B ahead
DecideBench agent_routing accuracy96.0 %40.0 %Decider 4B ahead
DecideBench content_moderation accuracy86.0 %38.0 %Decider 4B ahead
DecideBench action_review accuracy82.0 %38.0 %Decider 4B ahead
DecideBench support_intent accuracy98.0 %54.0 %Decider 4B ahead

Source · latest October 18, 2026

All data side by side

Decider 4BJulia 1
KindStandaloneClassifier
DeveloperMapikaSupersonic Labs
StatusOpen sourceOpen source
Weightsopenopen
Runs locallyyesyes
Parameters4 B0.14 B
LicenseApache-2.0Apache-2.0
Popularity (out of 100)4339