Compare / DeepSeek V3.2 Exp vs nanbeige4-3b-thinking-2511
DeepSeek V3.2 Exp vs nanbeige4-3b-thinking-2511
11 results from 2 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | DeepSeek V3.2 Exp | nanbeige4-3b-thinking-2511 |
|---|---|---|---|
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Irrelevance detection% correct | 93.2% | 83.1% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Memory% correct | 54.2% | 36.8% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Multi-turn tasks% correct | 44.9% | 51.1% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Overall accuracy% correct | 56.7% | 51.4% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Relevance detection% correct | 93.8% | 75.0% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (curated)% correct | 85.5% | 81.6% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (user-contributed)% correct | 76.0% | 79.4% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Web search% correct | 69.5% | 21.5% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 25.0% | 173.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.35 | 0.26 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 54.0 | 9.8 |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.