Ligue 1 Week 20 AI Model Audit: Top Performers and Surprises
Marin 8B Instruct led Ligue 1 predictions with 3.22 avg points/match, followed by Llama 3.2 3B Turbo (2.78) and Qwen 2.5 7B Turbo (2.56). Models achieved 38.86% correct tendency overall. Monaco vs Rennes (4-0) was a major upset, defying 76.9% consensus for Rennes.
Marin 8B Instruct led Ligue 1 predictions with 3.22 avg points/match, followed by Llama 3.2 3B Turbo (2.78) and Qwen 2.5 7B Turbo (2.56). Models achieved 38.86% correct tendency overall. Monaco vs Rennes (4-0) was a major upset, defying 76.9% consensus for Rennes.
Top 10 Models
| # | Model | Matches | Total Points | Avg Pts/Match | Tendency % | Exact % |
|---|---|---|---|---|---|---|
| 1 | Marin 8B Instruct (Marin Community) | 9 | 29 | 3.22 | 66.7% | 33.3% |
| 2 | Llama 3.2 3B Turbo (Meta) | 9 | 25 | 2.78 | 55.6% | 11.1% |
| 3 | Qwen 2.5 7B Turbo (Alibaba) | 9 | 23 | 2.56 | 44.4% | 22.2% |
| 4 | Qwen 2.5 72B Turbo (Alibaba) | 9 | 22 | 2.44 | 55.6% | 22.2% |
| 5 | Llama 3.1 8B Turbo (Meta) | 9 | 21 | 2.33 | 55.6% | 11.1% |
| 6 | Llama 3 8B Lite (Meta) | 9 | 21 | 2.33 | 44.4% | 33.3% |
| 7 | Kimi K2 0905 (Moonshot) | 9 | 20 | 2.22 | 44.4% | 11.1% |
| 8 | Llama 4 Maverick (Meta) | 9 | 19 | 2.11 | 44.4% | 22.2% |
| 9 | Llama 3.1 405B Turbo (Meta) | 9 | 19 | 2.11 | 44.4% | 22.2% |
| 10 | Gemma 3n E4B (Google) | 9 | 18 | 2.00 | 44.4% | 11.1% |
Match-by-Match Audit
- Strasbourg vs Paris Saint Germain: 88.5% correct tendency, 53.8% exact score hits
- Angers vs Metz: 44.0% correct tendency, 8.0% exact score hits
- Toulouse vs Auxerre: 19.2% correct tendency, 0.0% exact score hits
- Nice vs Stade Brestois 29: 60.0% correct tendency, 4.0% exact score hits
- Lyon vs Lille: 16.0% correct tendency, 0.0% exact score hits
- Monaco vs Rennes: 7.7% correct tendency, 0.0% exact score hits
- Lorient vs Nantes: 38.5% correct tendency, 38.5% exact score hits
- Paris FC vs Marseille: 24.0% correct tendency, 0.0% exact score hits
- Lens vs Le Havre: 51.9% correct tendency, 0.0% exact score hits
Biggest Consensus Misses
- Monaco vs Rennes (4-0): Consensus A (76.9%) | Counts H/D/A: 2/4/20
- Toulouse vs Auxerre (0-0): Consensus H (73.1%) | Counts H/D/A: 19/5/2
- Paris FC vs Marseille (2-2): Consensus A (72.0%) | Counts H/D/A: 1/6/18
- Lorient vs Nantes (2-1): Consensus D (61.5%) | Counts H/D/A: 10/16/0
- Lyon vs Lille (1-0): Consensus D (48.0%) | Counts H/D/A: 4/12/9
Methodology
Models were evaluated based on average points per match, with 6 points for exact score, 3 points for correct tendency, and 0 points otherwise.
Generation cost: $0.0022
Tokens: 4,444 input + 1,206 output
Frequently Asked Questions
What is this article about?
You might also like
Ligue 1 Week 23 AI Model Performance Audit
GLM-4.7 led Ligue 1 predictions with 3.22 points per match, followed by Qwen3 30B A3B (2.67) and MiniMax M2.5 (1.89). Models achieved 34.50% correct tendency overall, with Nantes' 2-0 win over Le Havre being the biggest consensus miss.
Feb 23, 2026
Ligue 1 AI Predictions Audit: Regular Season Week 22 Performance
Devstral Small and Llama 3.2 3B led Ligue 1 predictions this week with 3.00 points per match, followed by Phi-4 (2.75). Models achieved 40.68% correct tendency overall, though the Rennes 3-1 victory over Paris Saint Germain, predicted correctly by only 3.3% of models, was the biggest surprise.
Feb 16, 2026
Ligue 1 AI Predictions Audit - Regular Season 21
Llama 3.2 3B Turbo led Ligue 1 predictions with 2.89 avg points/match, followed by GPT-OSS 120B (2.17) and DeepSeek V3.1 Terminus (2.00). Models achieved 36.83% correct tendency overall. Biggest upset: Le Havre vs Strasbourg (2-1) with 66.7% consensus on draw.