CodeLlama-34B-Instruct-hf: benchmark results
Meta's CodeLlama 34B Instruct, the instruction-tuned code model in the Llama 2-based CodeLlama family, released under the Llama 2 Community License; the linked card provides no further details here. Provider: Meta. Released 2023-08-24. Access: Open.
Aggregate Rating 1352 ± 16, rank #1381 of 1656 rated models, from 44 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| ToolBench - Trip Booking | 89.17 | Task Score | 95.3 |
| ToolBench Leaderboard | 64.79 | Average Task Score | 95.3 |
| ToolBench - Google Sheets | 61.11 | Task Score | 93 |
| ToolBench - Home Search | 90 | Task Score | 93 |
| ToolBench - VirtualHome | 24.34 | Task Score | 93 |
| ToolBench - Tabletop | 47.46 | Task Score | 90.7 |
| HumanLikeness - Sound-1 | 78.53 | Humanlike Score (%) | 89.5 |
| ToolBench - WebShop Short | 25.99 | Task Score | 88.4 |
| HumanLikeness - Discourse-1 | 79.12 | Humanlike Score (%) | 84.2 |
| HumanLikeness - Sound-2 | 63.97 | Humanlike Score (%) | 84.2 |
| ToolBench - Open Weather | 94.28 | Task Score | 83.7 |
| HumanLikeness - Word-2 | 23.32 | Humanlike Score (%) | 78.9 |
Interactive version: theaggregate.ai/model?slug=codellama-34b-instruct-hf · How It Works · Data refreshed daily, snapshot 2026-10-20.