Skip to main content

MapikaJared PalmerDecision matchup

Decider 35B A3B NVFP4vs.Kev 4B

Which model makes better decisions?

Decider 35B A3B NVFP4 does better in independent tests, while Kev 4B is more popular.

Independent tests
Decider 35B A3B NVFP4ahead1 of 1
Kev 4B0 of 1
Jev Decision Index
Popularity
Decider 35B A3B NVFP40
Kev 4Bahead23
out of 100
Weights and access
Decider 35B A3B NVFP4open, local
Kev 4Bopen, local
Size
Decider 35B A3B NVFP435 B
Kev 4Bn/a
Parameters

Pick Decider 35B A3B NVFP4 for

Jev Decision Index

Pick Kev 4B for

more popular

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 35B A3B NVFP4 ahead

Decider 35B A3B NVFP4 is clearly ahead in 17 of 37 subtests and slightly in 11, Kev 4B in 2 clearly and slightly in 1. 6 are level.

POP909 · Decision Index 0.4.1

Decider 35B A3B NVFP46.7
Kev 4B1.2

2,000 tasks, as of October 28, 2026

BPoMP · Decision Index 0.4.1

Decider 35B A3B NVFP489.1
Kev 4B65.1

5,000 tasks, as of October 28, 2026

WinoGrande · Decision Index 0.4.1

Decider 35B A3B NVFP488.6
Kev 4B70.2

1,267 tasks, as of October 28, 2026

Show all 37 subtests
SubtestDecider 35B A3B NVFP4Kev 4BVerdict
POP909 · Decision Index 0.4.16.71.2Decider 35B A3B NVFP4 ahead
BPoMP · Decision Index 0.4.189.165.1Decider 35B A3B NVFP4 ahead
WinoGrande · Decision Index 0.4.188.670.2Decider 35B A3B NVFP4 ahead
CRUXEval · Decision Index 0.4.159.848.1Decider 35B A3B NVFP4 ahead
ChessBench · Decision Index 0.4.110.98.9Decider 35B A3B NVFP4 ahead
New Yorker · Decision Index 0.4.164.052.3Decider 35B A3B NVFP4 ahead
HellaSwag · Decision Index 0.4.196.778.9Decider 35B A3B NVFP4 ahead
PhishNChips · Decision Index 0.4.166.955.3Decider 35B A3B NVFP4 ahead
SATA-Bench · Decision Index 0.4.134.728.8Decider 35B A3B NVFP4 ahead
MMLU-Pro · Decision Index 0.4.160.550.7Decider 35B A3B NVFP4 ahead
BBH · Decision Index 0.4.171.362.5Decider 35B A3B NVFP4 ahead
Humicroedit · Decision Index 0.4.162.054.4Decider 35B A3B NVFP4 ahead
cfcolor · Decision Index 0.4.164.957.9Decider 35B A3B NVFP4 ahead
GSM8K · Decision Index 0.4.141.546.5Kev 4B ahead
When2Call · Decision Index 0.4.159.766.7Kev 4B ahead
Habermas · Decision Index 0.4.146.442.0Decider 35B A3B NVFP4 ahead
MuSR · Decision Index 0.4.161.856.3Decider 35B A3B NVFP4 ahead
HoVer · Decision Index 0.4.163.559.7Decider 35B A3B NVFP4 ahead
CLadder · Decision Index 0.4.166.763.1Decider 35B A3B NVFP4 ahead
ACOS · Decision Index 0.4.127.29.8Decider 35B A3B NVFP4 slightly ahead
ANLI · Decision Index 0.4.166.048.1Decider 35B A3B NVFP4 slightly ahead
Amazon ESCI · Decision Index 0.4.150.639.7Decider 35B A3B NVFP4 slightly ahead
CLINC150 · Decision Index 0.4.194.775.0Decider 35B A3B NVFP4 slightly ahead
iSarcasmEval · Decision Index 0.4.154.846.8Decider 35B A3B NVFP4 slightly ahead
ContractNLI · Decision Index 0.4.175.064.4Decider 35B A3B NVFP4 slightly ahead
ToolRet · Decision Index 0.4.166.361.0Decider 35B A3B NVFP4 slightly ahead
BANKING77 · Decision Index 0.4.189.882.9Decider 35B A3B NVFP4 slightly ahead
RAGTruth · Decision Index 0.4.163.167.5Kev 4B slightly ahead
BRIGHT · Decision Index 0.4.141.739.0Decider 35B A3B NVFP4 slightly ahead
NLI4CT · Decision Index 0.4.178.574.6Decider 35B A3B NVFP4 slightly ahead
VAST · Decision Index 0.4.154.552.4Decider 35B A3B NVFP4 slightly ahead
Home appliances · Decision Index 0.4.123.914.8level
HLE · Decision Index 0.4.18.810.2level
GPQA Diamond · Decision Index 0.4.140.837.8level
API-Bank · Decision Index 0.4.156.553.3level
FinEntity · Decision Index 0.4.185.287.1level
BFCL · Decision Index 0.4.192.293.1level

Source · latest October 28, 2026

All data side by side

Decider 35B A3B NVFP4Kev 4B
KindStandaloneAdapter
DeveloperMapikaJared Palmer
StatusOpen sourceOpen source
Releasedn/aSeptember 20, 2026
Weightsopenopen
Runs locallyyesyes
Parameters35 Bn/a
LicenseApache-2.0Apache-2.0 for the adapter and head
Popularity (out of 100)023