Gemini 3.1 Pro
Gemini 3.1 ProCompare inon
Benchmarks of Gemini 3.1 Pro5
BenchmarkByScoreRank
Coding2
Multimodal1
Legacy · saturated or no longer reported2
No benchmark matches
Rankingat high effort?
Alternatives?
ModelScoreCostPer taskWhat changes
OpenGPT OSS 20Bmedium-20%-99%$0.018 / $0.09Context: 1M → 131K+1
OpenQwQ 32B-24%-95%$0.15 / $0.40Context: 1M → 131K+1
OpenLlama 3.3 70B Instructunstated-28%-97%$0.10 / $0.32Context: 1M → 128K+1OpenLlama 3.1 70B Instructunstated-31%-91%$0.40 / $0.40Context: 1M → 131K+1mOpenReflection Llama 3.1 70B-34%-92%$0.35 / $0.40Context: 1M → 8K+1ModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT OSS 20BmediumOpen-20%-99%$0.018 / $0.09Context: 1M → 131K+1QwQ 32BOpen-24%-95%$0.15 / $0.40Context: 1M → 131K+1Llama 3.3 70B InstructunstatedOpen-28%-97%$0.10 / $0.32Context: 1M → 128K+1Llama 3.1 70B InstructunstatedOpen-31%-91%$0.40 / $0.40Context: 1M → 131K+1mReflection Llama 3.1 70BOpen-34%-92%$0.35 / $0.40Context: 1M → 8K+1ModelScoreCostPer taskWhat changes
Gemma 2 27BunstatedOpen-38%—$0.27 / $0.27Context: 1M → 8K+1Gemma 2 9B ItOpen-43%—$0.03 / $0.06Context: 1M → 8K+1Gemini 4 Argonhighn/a—*$1.99Synchronous output stops at 65,536+2Gemini 3.8 Flashmediumn/a—*$0.93Sampling parameters are ignored+1Gemini Robotics ER 2 Previewn/a——Context: not verifiedGemini 3.5 Flash Litehighn/a—$0.19Gemini 3.6 Flashhighn/a—*$1.60DiffusionGemma 26B A4BOpenn/a—$0.05 / $0.15Context: 1M → 256KAll Google models →Benchmarks
all categoriesChat & Knowledge
SimpleQA-Verifiedlab75.6
Chinese-SimpleQAlab85.9
GPQA Diamondlab94.3
MMLU-Pro EMlab91
HMMT February 2026lab94.7
IMOAnswerBenchlab81
Apexlab60.9
Apex Shortlistlab89.1
MRCR 1M MMRlab76.3
CorpusQA 1Mlab53.8
GDPval-AA Elolab1,314
HealthBench Professional length-adjustedlab43.8
Coding
SWE-bench Pro (edition unstated)lab54.2
SWE-bench Pro (public)lab54.2
DeepSWE (edition unstated)lab10
DeepSWE 1.1lab12
Terminal-Bench 2.1lab70.3
Terminal Bench 2.0lab68.5
LiveCodeBenchlab91.7
SWE Verifiedlab80.6
Codeforces Ratinglab3,052
Meta Internal Coding Benchlab59.2
Agents & Tool Use
OSWorld-Verifiedlab76.2
Toolathlon-Verifiedlab61.1
Toolathlonlab48.8
MCP Atlaslab78.2
BrowseComplab85.9
JobBenchlab15.9
Humanity’s Last Exam with toolslab51.4
Finance Agent v2lab43
HLE without toolslab44.4
HLE with toolslab51.6
MCPAtlas Publiclab69.2
DeepSearchQAlab71.3
Vision & Documents
Chartography26.1#19
CharXiv Reasoninglab81.6
BabyVisionlab51.5
FAQ
How much does Gemini 3.1 Pro cost?›
$2 per 1M input tokens and $12 per 1M output tokens (cheapest host: DeepInfra).
Is Gemini 3.1 Pro free to use?›
There is no free API; the cheapest paid price is $2 per 1M input tokens.
How good is Gemini 3.1 Pro, and what is it best at?›
#1 of 32 in Chat & Knowledge, #37 of 44 in Coding, #37 of 44 in Agents & Tool Use, #19 of 37 in Vision & Documents.
Is Gemini 3.1 Pro better than GPT-4o?›
On MMLU-Pro · official board, Gemini 3.1 Pro scores 91.2 vs 74.7 — higher.
How can I use Gemini 3.1 Pro?›
Through the API of DeepInfra.
What is the context window of Gemini 3.1 Pro?›
1M tokens.
Can Gemini 3.1 Pro read images?›
It accepts text and images, audio, video as input.
What is Gemini 3.1 Pro?›
Gemini 3.1 Pro is a language model from Google.
What are the best alternatives to Gemini 3.1 Pro?›
GPT OSS 20B (open weights).