Compare / llama-4-maverick-17b-128e-instruct-fp8 vs GPT-5.2
llama-4-maverick-17b-128e-instruct-fp8 vs GPT-5.2
10 results from 2 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | llama-4-maverick-17b-128e-instruct-fp8 | GPT-5.2 |
|---|---|---|---|
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Irrelevance detection% correct | 56.0% | 87.3% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Memory% correct | 18.9% | 45.8% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Multi-turn tasks% correct | 20.2% | 43.8% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Overall accuracy% correct | 37.3% | 55.9% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Relevance detection% correct | 100.0% | 75.0% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (curated)% correct | 88.7% | 81.8% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (user-contributed)% correct | 73.7% | 70.4% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Web search% correct | 28.0% | 75.5% |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 100.0% | 100.0% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 8.2% | 8.4% |
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.