MapikaDecision matchup
Decider 2Bvs.Decider 4B
Which model makes better decisions?
Decider 4B does better in independent tests, while Decider 2B is much smaller.
Pick Decider 2B for
Pick Decider 4B for
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 aheadDecider 4B is clearly ahead in 17 of 37 subtests and slightly in 11, Decider 2B in 1. 8 are level.
POP909 · Decision Index 0.4.1
2,000 tasks, as of October 28, 2026
Home appliances · Decision Index 0.4.1
88 tasks, as of October 28, 2026
API-Bank · Decision Index 0.4.1
508 tasks, as of October 28, 2026
Show all 37 subtests
| Subtest | Decider 2B | Decider 4B | Verdict |
|---|---|---|---|
| POP909 · Decision Index 0.4.1 | 0.3 | 10.0 | Decider 4B ahead |
| Home appliances · Decision Index 0.4.1 | 3.4 | 17.1 | Decider 4B ahead |
| API-Bank · Decision Index 0.4.1 | 31.7 | 60.4 | Decider 4B ahead |
| GSM8K · Decision Index 0.4.1 | 23.1 | 40.6 | Decider 4B ahead |
| New Yorker · Decision Index 0.4.1 | 44.7 | 62.9 | Decider 4B ahead |
| When2Call · Decision Index 0.4.1 | 54.7 | 75.9 | Decider 4B ahead |
| MMLU-Pro · Decision Index 0.4.1 | 37.6 | 51.7 | Decider 4B ahead |
| CRUXEval · Decision Index 0.4.1 | 33.0 | 43.7 | Decider 4B ahead |
| BPoMP · Decision Index 0.4.1 | 64.9 | 83.4 | Decider 4B ahead |
| SATA-Bench · Decision Index 0.4.1 | 24.7 | 31.6 | Decider 4B ahead |
| BBH · Decision Index 0.4.1 | 51.0 | 60.9 | Decider 4B ahead |
| WinoGrande · Decision Index 0.4.1 | 67.7 | 80.7 | Decider 4B ahead |
| PhishNChips · Decision Index 0.4.1 | 50.1 | 57.8 | Decider 4B ahead |
| CLadder · Decision Index 0.4.1 | 57.3 | 63.3 | Decider 4B ahead |
| HellaSwag · Decision Index 0.4.1 | 85.8 | 94.0 | Decider 4B ahead |
| cfcolor · Decision Index 0.4.1 | 57.0 | 60.5 | Decider 4B ahead |
| BFCL · Decision Index 0.4.1 | 82.9 | 88.0 | Decider 4B ahead |
| HoVer · Decision Index 0.4.1 | 59.1 | 56.3 | Decider 2B ahead |
| ACOS · Decision Index 0.4.1 | 7.9 | 19.2 | Decider 4B slightly ahead |
| iSarcasmEval · Decision Index 0.4.1 | 22.4 | 42.1 | Decider 4B slightly ahead |
| RAGTruth · Decision Index 0.4.1 | 28.3 | 52.2 | Decider 4B slightly ahead |
| BRIGHT · Decision Index 0.4.1 | 37.1 | 45.5 | Decider 4B slightly ahead |
| Amazon ESCI · Decision Index 0.4.1 | 34.7 | 42.4 | Decider 4B slightly ahead |
| CLINC150 · Decision Index 0.4.1 | 73.7 | 86.6 | Decider 4B slightly ahead |
| ANLI · Decision Index 0.4.1 | 49.4 | 56.4 | Decider 4B slightly ahead |
| ContractNLI · Decision Index 0.4.1 | 63.4 | 72.3 | Decider 4B slightly ahead |
| FinEntity · Decision Index 0.4.1 | 76.1 | 86.0 | Decider 4B slightly ahead |
| NLI4CT · Decision Index 0.4.1 | 65.5 | 73.7 | Decider 4B slightly ahead |
| VAST · Decision Index 0.4.1 | 51.0 | 53.5 | Decider 4B slightly ahead |
| HLE · Decision Index 0.4.1 | 12.0 | 10.4 | level |
| GPQA Diamond · Decision Index 0.4.1 | 36.2 | 41.3 | level |
| ChessBench · Decision Index 0.4.1 | 9.9 | 10.5 | level |
| Habermas · Decision Index 0.4.1 | 42.2 | 40.6 | level |
| ToolRet · Decision Index 0.4.1 | 63.1 | 65.0 | level |
| BANKING77 · Decision Index 0.4.1 | 72.8 | 74.7 | level |
| MuSR · Decision Index 0.4.1 | 54.9 | 53.7 | level |
| Humicroedit · Decision Index 0.4.1 | 59.9 | 59.7 | level |
Source · latest October 28, 2026
DecideBench
Decider 4B aheadDecider 4B is clearly ahead in 9 of 10 subtests, Decider 2B in none. 1 is level.
DecideBench claim_support accuracy
50 tasks, as of October 18, 2026
DecideBench pair accuracy
200 tasks, as of October 18, 2026
DecideBench returns_policy accuracy
50 tasks, as of October 18, 2026
Show all 10 subtests
| Subtest | Decider 2B | Decider 4B | Verdict |
|---|---|---|---|
| DecideBench claim_support accuracy | 42.0 % | 90.0 % | Decider 4B ahead |
| DecideBench pair accuracy | 41.0 % | 78.5 % | Decider 4B ahead |
| DecideBench returns_policy accuracy | 46.0 % | 82.0 % | Decider 4B ahead |
| DecideBench action_review accuracy | 56.0 % | 82.0 % | Decider 4B ahead |
| DecideBench agent_routing accuracy | 66.0 % | 96.0 % | Decider 4B ahead |
| DecideBench overall accuracy | 63.2 % | 88.5 % | Decider 4B ahead |
| DecideBench ticket_triage accuracy | 64.0 % | 86.0 % | Decider 4B ahead |
| DecideBench content_moderation accuracy | 66.0 % | 86.0 % | Decider 4B ahead |
| DecideBench review_sentiment accuracy | 70.0 % | 88.0 % | Decider 4B ahead |
| DecideBench support_intent accuracy | 96.0 % | 98.0 % | level |
Source · latest October 18, 2026
All data side by side
| Decider 2B | Decider 4B | |
|---|---|---|
| Kind | Standalone | Standalone |
| Developer | Mapika | Mapika |
| Status | Open source | Open source |
| Weights | open | open |
| Runs locally | yes | yes |
| Parameters | 1.9 B | 4 B |
| License | Apache-2.0 | Apache-2.0 |
| Popularity (out of 100) | 52 | 43 |