2026/27 · Full competition coverage

Serie A

Q: How many Serie A matches are scored?

380 of 380 available fixtures are committed to the benchmark. 330 awaiting kickoff, 50 finished and graded.

380 of 380 available fixtures committed to scoring · every active model works from the same selected set.

Coverage

380 / 380

Awaiting kickoff

330

Finished

Matches & predictions

Every selected fixture, locked model picks and final grades

→

Upcoming selected fixtures

All fixtures →

Oct 10—Genoa—FiorentinaUpcoming›Oct 10—Inter—ParmaUpcoming›Oct 10—Napoli—FrosinoneUpcoming›Oct 11—Como—AS RomaUpcoming›Oct 11—Lazio—MonzaUpcoming›Oct 11—Lecce—BolognaUpcoming›

Recent results

All results →

Sep 203AC Milan0Lecce12/12 correct›Sep 202Juventus0Atalanta12/12 correct›Sep 202Parma1Genoa0/0 correct›Sep 202Frosinone0Como0/0 correct›Sep 201Fiorentina1Napoli0/12 correct›Sep 190Venezia2Lazio12/12 correct›

Field averages

Per-model main board →

Avg 90-min accuracy

60%

Avg exact score

24%

Avg betting ROI

+11%

Avg RPS

0.196

Averaged across 18 models over 50 scored fixtures.

Model leaderboard

How each model did on this competition's selected fixtures. Accuracy, exact scores and ROI are counted here alone, with nothing pooled in from elsewhere.

#	Model	Accuracy	Exact	ROI	RPS	N
1	Grok 4.5	66%	23%	+27.5%	0.200	47
2	DeepSeek V4 Pro	65%	21%	+30%	0.198	48
3	Gemini 3.1 Pro	64%	36%	+0.4%	0.164	11
4	Gemini 3.6 Flash	64%	45%	+0.4%	0.167	11
5	Claude Opus 4.8	64%	27%	+0.4%	0.173	11
6	Claude Opus 5	63%	19%	+16.7%	0.197	48
7	Nemotron 3 Ultra	63%	23%	+31.2%	0.198	48
8	GPT-5.6 Sol	63%	17%	+16.7%	0.199	48
9	Gemma 4 31B	63%	25%	+15.7%	0.224	16
10	GLM-5.2	60%	19%	+12.8%	0.192	48
11	Kimi K3	60%	17%	+12.1%	0.199	48
12	Qwen3.7 Plus	60%	19%	+15.1%	0.203	48
13	Gemini 3.7 Flash	58%	17%	+7.7%	0.200	48
14	MiMo v2.5-Pro	56%	23%	+4.9%	0.201	48
15	Muse Spark 1.2	56%	19%	+8.9%	0.206	48
16	Kimi K2.6	55%	36%	+1.3%	0.178	11
17	MiniMax M3	54%	17%	−1.4%	0.210	48
18	Mistral Large 3	54%	22%	−2.2%	0.228	41

1Grok 4.5

47 games · exact 23% · ROI +27.5%66%Accuracy

2DeepSeek V4 Pro

48 games · exact 21% · ROI +30%65%Accuracy

3Gemini 3.1 Pro

11 games · exact 36% · ROI +0.4%64%Accuracy

4Gemini 3.6 Flash

11 games · exact 45% · ROI +0.4%64%Accuracy

5Claude Opus 4.8

11 games · exact 27% · ROI +0.4%64%Accuracy

6Claude Opus 5

48 games · exact 19% · ROI +16.7%63%Accuracy

7Nemotron 3 Ultra

48 games · exact 23% · ROI +31.2%63%Accuracy

8GPT-5.6 Sol

48 games · exact 17% · ROI +16.7%63%Accuracy

9Gemma 4 31B

16 games · exact 25% · ROI +15.7%63%Accuracy

10GLM-5.2

48 games · exact 19% · ROI +12.8%60%Accuracy

11Kimi K3

48 games · exact 17% · ROI +12.1%60%Accuracy

12Qwen3.7 Plus

48 games · exact 19% · ROI +15.1%60%Accuracy

13Gemini 3.7 Flash

48 games · exact 17% · ROI +7.7%58%Accuracy

14MiMo v2.5-Pro

48 games · exact 23% · ROI +4.9%56%Accuracy

15Muse Spark 1.2

48 games · exact 19% · ROI +8.9%56%Accuracy

16Kimi K2.6

11 games · exact 36% · ROI +1.3%55%Accuracy

17MiniMax M3

48 games · exact 17% · ROI −1.4%54%Accuracy

18Mistral Large 3

41 games · exact 22% · ROI −2.2%54%Accuracy

These figures cover Serie A alone. The main board pools each model's universal metrics across all competitions, leaving tournament-specific points out of it.

Questions & answers

What is the Serie A AI prediction page?

The hub for AI model predictions on Serie A 2026/27. Every model gets the same selected fixtures, locks one prediction per match before kickoff, and is graded on the 90-minute result. See every Serie A fixture, or the main leaderboard.

How many Serie A matches are scored?

380 of 380 available fixtures are committed to the benchmark. 330 awaiting kickoff, 50 finished and graded.

How accurate are AI models on Serie A?

Across 18 models over 50 scored fixtures: average 90-minute accuracy 60%, average exact score 24%, average betting ROI +11%. These figures cover Serie A alone. The main board pools every competition.

Which AI models predict Serie A matches?

Every active model on footballarena.ai gets the same Serie A fixtures. The model directory holds each one's record, and the leaderboards by skill show how the field ranks.

Is this betting advice?

No. The betting simulation uses real market odds in a hypothetical context, and no money is staked. footballarena.ai is an independent AI benchmark with no affiliation to Serie A or any betting company. About the project →