Reasoning / Benchmark profile
A suite of 23 challenging tasks from the BIG-Bench collaborative benchmark where prior language models failed to exceed average human performance, even with chain-of-thought prompting.
Observed results
5 ranked models · higher is better · labels show rank and score
| Model | Rank | Model ID | Provider | Score | Unit |
|---|---|---|---|---|---|
| DeepSeek V4 Pro Base | #1 | deepseek-v4-pro-base | DeepSeek | 87.5 | percent |
| DeepSeek V4 Flash Base | #2 | deepseek-v4-flash-base | DeepSeek | 86.9 | percent |
| Soofi S 30B-A3B | #3 | soofi-s-30b-a3b | Soofi Project | 78.8 | percent |
| MiniCPM5-1B | #4 | minicpm5-1b | OpenBMB | 71.89 | percent |
| Gemma 4 12B Unified | #5 | gemma-4-12b | 53 | percent |
One best compatible score per canonical product · higher is better
| Rank | Model | Weights | Evidence | Score |
|---|---|---|---|---|
| 01 | DeepSeek V4 Pro BaseDeepSeek | Open weights | Source carrier | 87.5% |
| 02 | DeepSeek V4 Flash BaseDeepSeek | Open weights | Source carrier | 86.9% |
| 03 | Soofi S 30B-A3BSoofi Project | Open weights | Source carrier | 78.8% |
| 04 | MiniCPM5-1BOpenBMB | Open weights | Source carrier | 71.9% |
| 05 | Gemma 4 12B UnifiedGoogle | Open weights | Source carrier | 53% |
Each distinct published measurement is retained. Same-snapshot canonical and configuration projections appear once.
| Published model / configuration | Score | Evidence & protocol | Source & dates |
|---|---|---|---|
Exact BenchLM registry variant DeepSeek V4 Pro Base; bulk export does not retain a complete upstream harness configuration.DeepSeek V4 Pro BaseExact identitySource label without a registered configuration ID Canonical product: deepseek-v4-pro-base | 87.5% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-08-01Observed Checked |
Exact BenchLM registry variant DeepSeek V4 Flash Base; bulk export does not retain a complete upstream harness configuration.DeepSeek V4 Flash BaseExact identitySource label without a registered configuration ID Canonical product: deepseek-v4-flash-base | 86.9% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-08-01Observed Checked |
Exact BenchLM registry variant Soofi S 30B-A3B; bulk export does not retain a complete upstream harness configuration.Soofi S 30B-A3BExact identitySource label without a registered configuration ID Canonical product: soofi-s-30b-a3b | 78.8% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-08-01Observed Checked |
Exact BenchLM registry variant MiniCPM5-1B; bulk export does not retain a complete upstream harness configuration.MiniCPM5-1BExact identitySource label without a registered configuration ID Canonical product: minicpm5-1b | 71.9% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-08-01Observed Checked |
Exact BenchLM registry variant Gemma 4 12B; bulk export does not retain a complete upstream harness configuration.Gemma 4 12B UnifiedExact identitySource label without a registered configuration ID Canonical product: gemma-4-12b | 53% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-08-01Observed Checked |
Exact BenchLM registry variant DeepSeek V4 Pro Base; bulk export does not retain a complete upstream harness configuration.DeepSeek V4 Pro BaseExact identitySource label without a registered configuration ID Canonical product: deepseek-v4-pro-base | 87.5% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-07-27Observed Checked |
Exact BenchLM registry variant DeepSeek V4 Flash Base; bulk export does not retain a complete upstream harness configuration.DeepSeek V4 Flash BaseExact identitySource label without a registered configuration ID Canonical product: deepseek-v4-flash-base | 86.9% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-07-27Observed Checked |
Exact BenchLM registry variant Soofi S 30B-A3B; bulk export does not retain a complete upstream harness configuration.Soofi S 30B-A3BExact identitySource label without a registered configuration ID Canonical product: soofi-s-30b-a3b | 78.8% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-07-27Observed Checked |
Exact BenchLM registry variant MiniCPM5-1B; bulk export does not retain a complete upstream harness configuration.MiniCPM5-1BExact identitySource label without a registered configuration ID Canonical product: minicpm5-1b | 71.9% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-07-27Observed Checked |
Exact BenchLM registry variant Gemma 4 12B; bulk export does not retain a complete upstream harness configuration.Gemma 4 12B UnifiedExact identitySource label without a registered configuration ID Canonical product: gemma-4-12b | 53% | Source carriersource-checkedVersion & system2022 Source-native system | BenchLM public datasets — 2026-07-27Observed Checked |
From result to context
A suite of 23 challenging tasks from the BIG-Bench collaborative benchmark where prior language models failed to exceed average human performance, even with chain-of-thought prompting.
The summary shows one best compatible source result per canonical product. Versions, effort settings and execution systems remain attached to the underlying records.
Scores from different versions or harnesses may not be interchangeable. The published source rows preserve those distinctions and their original units. A source result is not automatically an input to a current index.
This catalogue entry preserves available source evidence. Current indices admit only their specifically reviewed tracks and configurations.
Contamination risk: Unknown. Lifecycle: Active.
