Jared PalmerInternLMDecision matchup
Kev 9Bvs.Intern-Decision-4B
Which model makes better decisions?
Kev 9B does better in independent tests.
Independent testsJev Decision Index
Kev 9Bahead1 of 1
Intern-Decision-4B0 of 1
Popularityout of 100
Kev 9B40
Intern-Decision-4B40
Weights and access
Kev 9Bopen, local
Intern-Decision-4Bopen, local
SizeParameters
Kev 9Bn/a
Intern-Decision-4B4 B
Pick Kev 9B for
Jev Decision Index
Pick Intern-Decision-4B for
No clear edge
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
Kev 9B aheadKev 9B is clearly ahead in 10 of 37 subtests and slightly in 9, Intern-Decision-4B in 4 clearly and slightly in 2. 12 are level.
GSM8K · Decision Index 0.4.1
Kev 9B54.4
Intern-Decision-4B25.6
2,638 tasks, as of October 28, 2026
POP909 · Decision Index 0.4.1
Kev 9B10.2
Intern-Decision-4B7.0
2,000 tasks, as of October 28, 2026
BPoMP · Decision Index 0.4.1
Kev 9B58.9
Intern-Decision-4B81.8
5,000 tasks, as of October 28, 2026
Show all 37 subtests
| Subtest | Kev 9B | Intern-Decision-4B | Verdict |
|---|---|---|---|
| GSM8K · Decision Index 0.4.1 | 54.4 | 25.6 | Kev 9B ahead |
| POP909 · Decision Index 0.4.1 | 10.2 | 7.0 | Kev 9B ahead |
| BPoMP · Decision Index 0.4.1 | 58.9 | 81.8 | Intern-Decision-4B ahead |
| CRUXEval · Decision Index 0.4.1 | 55.4 | 45.1 | Kev 9B ahead |
| MMLU-Pro · Decision Index 0.4.1 | 56.5 | 46.1 | Kev 9B ahead |
| API-Bank · Decision Index 0.4.1 | 65.4 | 76.0 | Intern-Decision-4B ahead |
| Habermas · Decision Index 0.4.1 | 40.0 | 45.1 | Intern-Decision-4B ahead |
| cfcolor · Decision Index 0.4.1 | 53.0 | 58.5 | Intern-Decision-4B ahead |
| WinoGrande · Decision Index 0.4.1 | 75.1 | 68.4 | Kev 9B ahead |
| BBH · Decision Index 0.4.1 | 66.3 | 62.1 | Kev 9B ahead |
| HellaSwag · Decision Index 0.4.1 | 82.7 | 77.6 | Kev 9B ahead |
| CLadder · Decision Index 0.4.1 | 63.5 | 60.0 | Kev 9B ahead |
| When2Call · Decision Index 0.4.1 | 60.7 | 57.8 | Kev 9B ahead |
| BFCL · Decision Index 0.4.1 | 95.2 | 92.7 | Kev 9B ahead |
| VAST · Decision Index 0.4.1 | 56.7 | 40.7 | Kev 9B slightly ahead |
| BANKING77 · Decision Index 0.4.1 | 83.3 | 62.6 | Kev 9B slightly ahead |
| ToolRet · Decision Index 0.4.1 | 63.8 | 52.8 | Kev 9B slightly ahead |
| BRIGHT · Decision Index 0.4.1 | 38.2 | 33.1 | Kev 9B slightly ahead |
| FinEntity · Decision Index 0.4.1 | 89.7 | 80.9 | Kev 9B slightly ahead |
| Amazon ESCI · Decision Index 0.4.1 | 49.1 | 45.3 | Kev 9B slightly ahead |
| iSarcasmEval · Decision Index 0.4.1 | 43.2 | 46.4 | Intern-Decision-4B slightly ahead |
| ACOS · Decision Index 0.4.1 | 17.0 | 15.9 | Kev 9B slightly ahead |
| CLINC150 · Decision Index 0.4.1 | 77.1 | 72.7 | Kev 9B slightly ahead |
| ContractNLI · Decision Index 0.4.1 | 65.3 | 62.0 | Kev 9B slightly ahead |
| NLI4CT · Decision Index 0.4.1 | 75.2 | 77.8 | Intern-Decision-4B slightly ahead |
| HLE · Decision Index 0.4.1 | 11.0 | 12.8 | level |
| Home appliances · Decision Index 0.4.1 | 11.4 | 12.5 | level |
| GPQA Diamond · Decision Index 0.4.1 | 37.8 | 41.3 | level |
| MuSR · Decision Index 0.4.1 | 56.5 | 53.7 | level |
| ChessBench · Decision Index 0.4.1 | 12.0 | 12.6 | level |
| New Yorker · Decision Index 0.4.1 | 56.6 | 58.5 | level |
| Humicroedit · Decision Index 0.4.1 | 57.3 | 58.6 | level |
| RAGTruth · Decision Index 0.4.1 | 70.6 | 71.9 | level |
| ANLI · Decision Index 0.4.1 | 58.1 | 58.4 | level |
| SATA-Bench · Decision Index 0.4.1 | 30.7 | 30.8 | level |
| HoVer · Decision Index 0.4.1 | 60.7 | 60.5 | level |
| PhishNChips · Decision Index 0.4.1 | 61.6 | 61.5 | level |
Source · latest October 28, 2026
All data side by side
| Kev 9B | Intern-Decision-4B | |
|---|---|---|
| Kind | Adapter | Standalone |
| Developer | Jared Palmer | InternLM |
| Status | Open source | Open source |
| Released | September 20, 2026 | n/a |
| Weights | open | open |
| Runs locally | yes | yes |
| Parameters | n/a | 4 B |
| License | Apache-2.0 for the adapter and head | Apache-2.0 with the Qwen notices retained |
| Popularity (out of 100) | 40 | 40 |