CloudflareInternLMDecision matchup
Clef-Flashvs.Intern-Decision-4B
Which model makes better decisions?
Clef-Flash does better in independent tests, while Intern-Decision-4B is much smaller.
Pick Clef-Flash for
Pick Intern-Decision-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
Clef-Flash aheadClef-Flash is clearly ahead in 21 of 37 subtests and slightly in 8, Intern-Decision-4B in 1 clearly and slightly in 3. 4 are level.
Home appliances · Decision Index 0.4.1
88 tasks, as of October 28, 2026
POP909 · Decision Index 0.4.1
2,000 tasks, as of October 28, 2026
CRUXEval · Decision Index 0.4.1
570 tasks, as of October 28, 2026
Show all 37 subtests
| Subtest | Clef-Flash | Intern-Decision-4B | Verdict |
|---|---|---|---|
| Home appliances · Decision Index 0.4.1 | 97.7 | 12.5 | Clef-Flash ahead |
| POP909 · Decision Index 0.4.1 | 1.6 | 7.0 | Intern-Decision-4B ahead |
| CRUXEval · Decision Index 0.4.1 | 86.5 | 45.1 | Clef-Flash ahead |
| ChessBench · Decision Index 0.4.1 | 23.1 | 12.6 | Clef-Flash ahead |
| GSM8K · Decision Index 0.4.1 | 46.7 | 25.6 | Clef-Flash ahead |
| CLadder · Decision Index 0.4.1 | 97.7 | 60.0 | Clef-Flash ahead |
| MuSR · Decision Index 0.4.1 | 86.2 | 53.7 | Clef-Flash ahead |
| Habermas · Decision Index 0.4.1 | 71.8 | 45.1 | Clef-Flash ahead |
| WinoGrande · Decision Index 0.4.1 | 97.6 | 68.4 | Clef-Flash ahead |
| MMLU-Pro · Decision Index 0.4.1 | 65.2 | 46.1 | Clef-Flash ahead |
| Humicroedit · Decision Index 0.4.1 | 75.1 | 58.6 | Clef-Flash ahead |
| HellaSwag · Decision Index 0.4.1 | 98.6 | 77.6 | Clef-Flash ahead |
| GPQA Diamond · Decision Index 0.4.1 | 51.5 | 41.3 | Clef-Flash ahead |
| API-Bank · Decision Index 0.4.1 | 93.1 | 76.0 | Clef-Flash ahead |
| PhishNChips · Decision Index 0.4.1 | 75.1 | 61.5 | Clef-Flash ahead |
| SATA-Bench · Decision Index 0.4.1 | 36.6 | 30.8 | Clef-Flash ahead |
| BPoMP · Decision Index 0.4.1 | 95.4 | 81.8 | Clef-Flash ahead |
| New Yorker · Decision Index 0.4.1 | 66.3 | 58.5 | Clef-Flash ahead |
| When2Call · Decision Index 0.4.1 | 65.3 | 57.8 | Clef-Flash ahead |
| cfcolor · Decision Index 0.4.1 | 65.9 | 58.5 | Clef-Flash ahead |
| BBH · Decision Index 0.4.1 | 68.9 | 62.1 | Clef-Flash ahead |
| BFCL · Decision Index 0.4.1 | 98.8 | 92.7 | Clef-Flash ahead |
| iSarcasmEval · Decision Index 0.4.1 | 19.7 | 46.4 | Intern-Decision-4B slightly ahead |
| RAGTruth · Decision Index 0.4.1 | 35.6 | 71.9 | Intern-Decision-4B slightly ahead |
| ACOS · Decision Index 0.4.1 | 26.0 | 15.9 | Clef-Flash slightly ahead |
| BANKING77 · Decision Index 0.4.1 | 90.8 | 62.6 | Clef-Flash slightly ahead |
| ContractNLI · Decision Index 0.4.1 | 84.5 | 62.0 | Clef-Flash slightly ahead |
| Amazon ESCI · Decision Index 0.4.1 | 57.4 | 45.3 | Clef-Flash slightly ahead |
| ToolRet · Decision Index 0.4.1 | 66.2 | 52.8 | Clef-Flash slightly ahead |
| VAST · Decision Index 0.4.1 | 49.6 | 40.7 | Clef-Flash slightly ahead |
| FinEntity · Decision Index 0.4.1 | 97.1 | 80.9 | Clef-Flash slightly ahead |
| BRIGHT · Decision Index 0.4.1 | 38.3 | 33.1 | Clef-Flash slightly ahead |
| CLINC150 · Decision Index 0.4.1 | 66.9 | 72.7 | Intern-Decision-4B slightly ahead |
| HLE · Decision Index 0.4.1 | 11.8 | 12.8 | level |
| ANLI · Decision Index 0.4.1 | 59.1 | 58.4 | level |
| HoVer · Decision Index 0.4.1 | 61.2 | 60.5 | level |
| NLI4CT · Decision Index 0.4.1 | 78.6 | 77.8 | level |
Source · latest October 28, 2026
Zero-shot IE benchmark
levelZero-shot IE spectrum: mixed-pool classification accuracy
48 tasks, as of October 17, 2026
Source · latest October 17, 2026
All data side by side
| Clef-Flash | Intern-Decision-4B | |
|---|---|---|
| Kind | Standalone | Standalone |
| Developer | Cloudflare | InternLM |
| Status | Active | Open source |
| Released | October 1, 2026 | n/a |
| Weights | open | open |
| Runs locally | yes | yes |
| Parameters | 9 B | 4 B |
| License | Apache-2.0 | Apache-2.0 with the Qwen notices retained |
| Price | $0.090 per 1M input tokens (8,182 neurons per 1M) | n/a |
| Popularity (out of 100) | 51 | 40 |