Gemma 4 31B Instruct (Venice): real-world results
Gemma 4 31B Instruct (Venice) passed 36 of 42 runs (86%) on 42 different use cases.
Average cost per run was $0.0070, average wall time 5m, with 0.10 human step-ins per run.
Results by category
| Category | Runs | Passed |
|---|---|---|
| Software & DevOps | 7 | 7/7 |
| Games & Simulations | 4 | 4/4 |
| Sales, Marketing & Support | 3 | 3/3 |
| Travel & Bookings | 3 | 3/3 |
| Finance / Retirement Planning | 2 | 2/2 |
| Analytics / Metrics | 1 | 0/1 |
| Aviation / Crew Scheduling Compliance | 1 | 0/1 |
| Content & Creative | 1 | 1/1 |
| Finance / Backtesting | 1 | 1/1 |
| Finance / Portfolio | 1 | 0/1 |
| Finance / Retirement Tax | 1 | 1/1 |
| Finance / Statement Analysis | 1 | 1/1 |
| Finance / Tax | 1 | 1/1 |
| Finance / Valuation | 1 | 1/1 |
| IT / DNS Administration | 1 | 0/1 |
| Life Admin & Home | 1 | 1/1 |
| Marine / Fishing Compliance | 1 | 0/1 |
| Marine / Fishing Planning | 1 | 1/1 |
| Marine / Powerboat Planning | 1 | 1/1 |
| Marine / Sailing Performance | 1 | 0/1 |
| Marine / Weather Routing | 1 | 1/1 |
| Market Research / Survey Analysis | 1 | 1/1 |
| Marketing / Data Cleaning | 1 | 1/1 |
| Security, Science & Hardware | 1 | 1/1 |
| Shopping & Errands | 1 | 1/1 |
| Software / Version Control | 1 | 1/1 |
| Strategy / Portfolio Prioritisation | 1 | 1/1 |
| Work & Operations | 1 | 1/1 |
Head-to-head comparisons
- Gemma 4 31B Instruct (Venice) vs Claude Haiku 5.5
- Gemma 4 31B Instruct (Venice) vs Claude Opus 5.5
- Gemma 4 31B Instruct (Venice) vs Claude Sonnet 5.5
- Gemma 4 31B Instruct (Venice) vs DeepSeek V4.1 Flash (Venice)
- Gemma 4 31B Instruct (Venice) vs Gemini 3.8 Flash
- Gemma 4 31B Instruct (Venice) vs Gemini 4 Argon
- Gemma 4 31B Instruct (Venice) vs GPT-6.1 Sol
- Gemma 4 31B Instruct (Venice) vs Grok 4.3 (Venice)
- Gemma 4 31B Instruct (Venice) vs Kimi K3.1 Pro (Venice)
- Gemma 4 31B Instruct (Venice) vs Llama 3.2 3B (local)
- Gemma 4 31B Instruct (Venice) vs MiniMax M2.5 (Venice)
- Gemma 4 31B Instruct (Venice) vs Mistral Small 3.2 24B Instruct (Venice)
- Gemma 4 31B Instruct (Venice) vs Qwen2.5 7B (local)
- Gemma 4 31B Instruct (Venice) vs Qwen3 Coder 480B Turbo (Venice)
- Gemma 4 31B Instruct (Venice) vs GLM 5.3 (Venice)
- Gemma 4 31B Instruct (Venice) vs GLM 4.7 (Venice)
Browse every run by Gemma 4 31B Instruct (Venice) →
From published single-model runs only (all settings grouped). How we score.