Model breakdown
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
3rd placeStill on the podium
Profile
Llama 3.2 11B VisionMeta
Release
2024-09-25Available on OpenRouter
Specs
11B131,072 tokens
Capabilities
Text + ImageOpen multimodal vision-language understanding and image Q&A
Training
Not publicly disclosedMeta AI
Rank#3
-234.5alignment score
84.1%crowd match
Mean gap15.9%
Human match84.1%
Best fitAvocado Tea
Average vote61.0%
61.0%model yes
62.8%human yes
Workload2K evals
2Kevals
100iterations
9.3Mtokens
Photo-by-photo

Model Results

Breaking down how close the model answered each question, compared to humans.

Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
90.0% yes10.0% no
Gap6.3%
Model readLeans yes
Dodge Van
Photo 02Dodge Van
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
11.0% yes89.0% no
Gap4.0%
Model readLeans no
Sub Sandwich
Photo 03Sub Sandwich
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
85.0% yes15.0% no
Gap9.5%
Model readLeans yes
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
38.0% yes62.0% no
Gap2.9%
Model readLeans no
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
89.0% yes11.0% no
Gap6.6%
Model readLeans yes
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
27.0% yes73.0% no
Gap27.2%
Model readLeans no
Hamburger
Photo 08Hamburger
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
85.0% yes15.0% no
Gap12.0%
Model readLeans yes
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
79.0% yes21.0% no
Gap19.6%
Model readLeans yes
Hot Dog
Photo 10Hot Dog
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
42.0% yes58.0% no
Gap2.2%
Model readLeans no
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
95.0% yes5.0% no
Gap29.4%
Model readLeans yes
Avocado Tea
Photo 12Avocado Tea
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
95.0% yes5.0% no
Gap2.2%
Model readLeans yes
Panini
Photo 13Panini
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
78.0% yes22.0% no
Gap14.4%
Model readLeans yes
Cookie PB
Photo 14Cookie PB
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
7.0% yes93.0% no
Gap44.5%
Model readLeans no
Chicken Wrap
Photo 15Chicken Wrap
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
52.0% yes48.0% no
Gap29.4%
Model readLeans yes
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
29.0% yes71.0% no
Gap37.3%
Model readLeans no
Sloppy Joe
Photo 17Sloppy Joe
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
77.0% yes23.0% no
Gap2.4%
Model readLeans yes
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
25.0% yes75.0% no
Gap4.8%
Model readLeans no
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
47.0% yes53.0% no
Gap8.7%
Model readLeans no
Bagel PB&J
Photo 20Bagel PB&J
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
89.0% yes11.0% no
Gap42.4%
Model readLeans yes
PhotoVote SplitHuman responseGapRead
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
90.0% yes10.0% no
6.3%absolute gap
Leans yesPeople mostly said yes
Dodge Van
Photo 02Dodge Van
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
11.0% yes89.0% no
4.0%absolute gap
Leans noPeople mostly said no
Sub Sandwich
Photo 03Sub Sandwich
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
85.0% yes15.0% no
9.5%absolute gap
Leans yesPeople mostly said yes
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
38.0% yes62.0% no
2.9%absolute gap
Leans noHuman knife-edge
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
89.0% yes11.0% no
6.6%absolute gap
Leans yesPeople mostly said yes
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
79.0% yes21.0% no
12.7%absolute gap
Leans yesPeople mostly said yes
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
27.0% yes73.0% no
27.2%absolute gap
Leans noHuman knife-edge
Hamburger
Photo 08Hamburger
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
85.0% yes15.0% no
12.0%absolute gap
Leans yesSplit concept
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
79.0% yes21.0% no
19.6%absolute gap
Leans yesHuman knife-edge
Hot Dog
Photo 10Hot Dog
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
42.0% yes58.0% no
2.2%absolute gap
Leans noSplit concept
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
95.0% yes5.0% no
29.4%absolute gap
Leans yesSplit concept
Avocado Tea
Photo 12Avocado Tea
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
95.0% yes5.0% no
2.2%absolute gap
Leans yesPeople mostly said yes
Panini
Photo 13Panini
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
78.0% yes22.0% no
14.4%absolute gap
Leans yesPeople mostly said yes
Cookie PB
Photo 14Cookie PB
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
7.0% yes93.0% no
44.5%absolute gap
Leans noHuman knife-edge
Chicken Wrap
Photo 15Chicken Wrap
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
52.0% yes48.0% no
29.4%absolute gap
Leans yesSplit concept
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
29.0% yes71.0% no
37.3%absolute gap
Leans noSplit concept
Sloppy Joe
Photo 17Sloppy Joe
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
77.0% yes23.0% no
2.4%absolute gap
Leans yesSplit concept
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
25.0% yes75.0% no
4.8%absolute gap
Leans noSplit concept
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
47.0% yes53.0% no
8.7%absolute gap
Leans noHuman knife-edge
Bagel PB&J
Photo 20Bagel PB&J
Meta / Llamameta-llama/llama-3.2-11b-vision-instruct
89.0% yes11.0% no
42.4%absolute gap
Leans yesHuman knife-edge
meta-llama/llama-3.2-11b-vision-instruct Sandwich Benchmark Breakdown | opensandwich.ai