Molmo2 8B
Dec 2025
Molmo2 8BCompare inon
Benchmarks of Molmo2 8B17
BenchmarkByScoreRank
Overall1
Knowledge9
Science2
Coding1
Multimodal2
Legacy · saturated or no longer reported2
No benchmark matches
—per task?
self-host onlyprice
—output speed?
37Kcontext
Rankingin non-thinking mode?
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-5 Minihighwins 62%—$0.053
OpenGemma 3 12Bunstatedwins 56%—$0.05 / $0.10Memory (4-bit): 4.3 GB → 6.1 GBBetterGPT-6 Lunamaxwins 65%—$0.068Function calling only with none · Chat Completions+3BetterGPT-5.6 Lunaxhighwins 69%—$0.085BetterGPT-5.6 Terraxhighwins 72%—$0.63Prompts with >272K input tokensModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-5 Minihighwins 62%—$0.053GPT-6 Lunamaxwins 65%—$0.068Function calling only with none · Chat Completions+3GPT-5.6 Lunaxhighwins 69%—$0.085GPT-5.6 Terraxhighwins 72%—$0.63Prompts with >272K input tokensGPT-6.1 Solmaxwins 72%—$0.72Remove unsupported parameters+5ModelScoreCostPer taskWhat changes
Gemma 3 12BunstatedOpenwins 56%—$0.05 / $0.10Memory (4-bit): 4.3 GB → 6.1 GB
Granite 4.2 8BunstatedOpenwins 54%—$0.06 / $0.25Memory (4-bit): 4.3 GB → 4.4 GB
NVIDIA Nemotron 3.5 Lightning 30B A3B NVFP4Openwins 53%——Memory (4-bit): 4.3 GB → 11 GBGemma 3n E4B InstructunstatedOpensame——Context: 37K → 32KGemma 3 4BunstatedOpensame—$0.04 / $0.05ModelScoreCostPer taskWhat changes
Olmo 3.1 32B InstructunstatedOpensame—$0.20 / $0.60Memory (4-bit): 4.3 GB → 19 GBOlmo 3 32B ThinkunstatedOpenwins 47%——Memory (4-bit): 4.3 GB → 19 GBOlmo 3.1 32B ThinkunstatedOpenwins 41%——Memory (4-bit): 4.3 GB → 19 GBOlmo 3 7B InstructOpenn/a——Olmo 3 7B ThinkOpenn/a——All Allen Institute for AI models →Run it yourself
estimateFits on✓ 8 GB (phone, laptop) · compressed✓ 16 GB · 8-bit✓ 24 GB GPU · full quality✓ 48 GB GPU · full quality✓ 80 GB GPU · full quality
LicenceApache-2.0free for commercial use
Benchmarks
all categoriesChat & Knowledge
Humanity's Last Exam4.3#301
AA Omniscience Index-69.2#291
CritPt (physics)0#180
AA Omniscience · Accuracy11.5#271
GPQA Diamond42.5#276
Coding
Agents & Tool Use
τ²-bench · AA0#232
Writing & Language
IFBench (instructions)26.9#221
Vision & Documents
MMMU-Pro (vision)37.5#124
FAQ
How much does Molmo2 8B cost?›
There is no per-unit API price: self-host only.
Is Molmo2 8B free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Molmo2 8B, and what is it best at?›
#298 of 334 in Chat & Knowledge, #210 of 239 in Coding, #229 of 242 in Agents & Tool Use, #218 of 245 in Writing & Language, #121 of 124 in Vision & Documents.
Is Molmo2 8B better than GPT-4o?›
On Arena Text · non-English, Molmo2 8B scores 1,276.2 vs 1,282.8 — about the same.
What is the context window of Molmo2 8B?›
37K tokens.
When was Molmo2 8B released? Is there a newer version?›
Released 2025-12-11. It is the current version in its line.
Is Molmo2 8B open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Molmo2 8B locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Molmo2 8B read images?›
It accepts text and video as input.
What is Molmo2 8B?›
Molmo2 8B is an open-weights language model from Allen Institute for AI, released Dec 2025.
What are the best alternatives to Molmo2 8B?›
GPT-5 Mini (better), Gemma 3 12B (open weights).