Skip to content

Google · Chat

Gemma 3 12B IT

Google 🇺🇸ChatGemini familyOpen weights. One of its sources says the weights are publicly available.The maker currently plans to deprecate it within the next 180 days (30 Mar 2027). Plans can change.

Newer version: Gemma 4 12B It. We compared release dates and version names in our catalogue.

From Aiml API's listing:
Gemma 3 12B is a mid-size open language model from Google offering a balance of capability and efficiency.
llminfo IQ
107.6
Market median price
$0.0794 blended / 1M
Max context
131K
Listings
31

Evidence and sources

🤗 google/gemma-3-12b-itllminfo IQ 107.6, ranked among chat models. From 331 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 68.1, 95% interval 49–87. From 10 benchmark families across 7 sources. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedPoor value (41.7). Costs more per task than its benchmark results predict. From 3 benchmarks with measured task costs. A fair amount of data behind it.How Value for money worksReleased 10 Mar 2025 Training-data cutoff, as its sources state it Maker HQ: US (according to Google's own website)
All rankings: #205 on OpenRouter this week #217 on OpenRouter this month #223 on OpenRouter todaySources: LMArena data, LiteLLM data, Models.dev data, AI/ML API data, OpenRouter data docs, llminfo researchPrimary data from LMArena from LiteLLM: deprecation date from Models.dev: weights availability from AI/ML API: primary mode, description from OpenRouter: knowledge cutoff from llminfo research: release date

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Advantages

Open weights available

The weights are public. 4 of the 8 peers that say either way have open weights.

Trade-offs

Deprecation scheduled

The maker plans to deprecate it on 30 Mar 2027, within the next 180 days.

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 04:34 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 6 maker's prices; 4 market median prices.

Input & output

as stated by the model's providers
Text
Images
Text

Capabilities

Supported operations and provider-stated features

VisionReasoningFunction callingStructured outputPrompt cachingBatch API

Listings

 Flowdotgoogle/gemma-3-12b-it:free · freeSource: FlowDot data33K$0.00$0.00
 OpenRoutergoogle/gemma-3-12b-it:free · freeSource: Helicone Costs dataNot published by the provider$0.00$0.00
 Kilogoogle/gemma-3-12b-itSources: Kilo AI Gateway data, Models.dev data131K$0.04$0.12
 Eden AIInference by DeepInfradeepinfra/google/gemma-3-12b-itSource: Models.dev data131K$0.05$0.15
 Fastroutergoogle/gemma-3-12b-itSource: FastRouter data131K$0.05$0.15
 Hugging FaceInference by DeepInfragoogle/gemma-3-12b-itSource: Hugging Face Inference Providers data131K$0.05$0.15
 Nous Portalgoogle/gemma-3-12b-itSource: Nous Portal Inference Gateway data131K$0.05$0.15
 DeepInfragoogle/gemma-3-12b-itSources: DeepInfra data, LiteLLM data, Models.dev data131K$0.05$0.15
 Novita AIgoogle/gemma-3-12b-itSources: LiteLLM data, Models.dev data, Novita AI data131K$0.05$0.10
 Aimlapigoogle/gemma-3-12b-itSource: Models.dev data131K$0.0529cache read $0.0423$0.1587
Show all 28 listings

Showing 10 of 28: the cheapest per 1M input tokens from each provider.

Also sold by Aihubmix, Amazon Web Services, Anonrouter, Crusoe, Databricks, Merge Gateway, Neon, Nextbit.

3 listings without a price are hidden.

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Multilingual

ECLeKTicSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
IndicParam42 related benchmarks · 42 results
Show 36 more resultsHide 36 results
SeaLLMs16 related benchmarks · 16 results
Show 10 more resultsHide 10 results
WMT24++Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Knowledge

Human preference

OCR

OmniDocBenchSource: Nanonets IDP data
TextVQASource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Coding

HumanEvalSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
MBPPSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Math

GSM8KSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
MATHSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Vision

AI2DSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
ChartQASource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
DocVQASource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
InfoVQASource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

General

Reasoning

Science

GPQASource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Factuality

SimpleQASource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Instruction following

IFEvalSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Safety

PhareSource: Phare

Agentic

Other

ArmBench-LLM Armenian22 related benchmarks · 22 results
Horangi36 related benchmarks · 36 results
Nejumi 428 related benchmarks · 28 results

Data credits: Data by LLM Stats · SEA-HELM leaderboard, AI Singapore (AI Products team), National University of Singapore. Published aggregate scores; results licence not stated (SEA-HELM code is MIT). · LMArena, CC BY 4.0 (licence) · ArmBench-LLM by Metric-AI. Public numeric primary task results; no separate results licence asserted. · ArtifactsBench (Tencent Hunyuan), CC BY 4.0 (licence) · EQ-Bench (Samuel J. Paech) — cite paech2023eqbench · Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · OpenRouter public benchmark leaderboard and search-lane exports, CC BY 4.0. Fractional source scores converted to percentages; immutable dated snapshots. (licence) · Horangi 4, Weights & Biases Korea; public publisher-selected numeric evaluation results · Speakleash community evaluation of INCLUDE-base-44 (Romanou et al., arXiv:2411.19799). Public results.csv; no results licence stated. · Maheshwari et al., IndicParam, arXiv:2512.00333v3. Cited numeric scores under the owner's 2026-10-10 task-only restriction scope decision; no results licence established. Tasks, prompts and generations excluded. · LLM Chess by Maxim Saplin, Apache-2.0 (Copyright [2024] [Maxim Saplin]) (licence) · Hugging Face official benchmark leaderboards; licence varies by benchmark · Nejumi Leaderboard 4, Weights & Biases Japan; public publisher-selected numeric evaluation results · Phare (Giskard), CC BY 4.0 (licence) · SeaExam and SeaBench leaderboard, SeaLLMs (DAMO-NLP-SG). Published leaderboard scores; results licence not stated. · Liashkov et al., Soro: A Lightweight Foundation Model and Chatbot for Tajik, arXiv:2605.27379v2, CC BY 4.0. Table 6 averages. (licence) · Swallow LLM Leaderboard v2, Swallow LLM team (Institute of Science Tokyo). Contents CC BY 4.0; Japanese post-trained task scores on the source's 0..1 scale. (licence) · UzLiB, Tahrirchi (Abror Shopulatov). Public LEADERBOARD.md numbers; repository MIT covers software and documentation, no separate results licence stated and none asserted. · Vectara Hallucination Leaderboard, Apache-2.0 (licence)

Also known as: amazon/google.gemma-3-12b-it, amazon/google.gemma-3-12b-it@us, bedrock/ap-northeast-1/google.gemma-3-12b-it, bedrock/ap-south-1/google.gemma-3-12b-it, bedrock/eu-south-1/google.gemma-3-12b-it, bedrock/eu-west-1/google.gemma-3-12b-it, bedrock/eu-west-2/google.gemma-3-12b-it, bedrock/sa-east-1/google.gemma-3-12b-it, bedrock_mantle/ap-northeast-1/google.gemma-3-12b-it, bedrock_mantle/ap-south-1/google.gemma-3-12b-it, bedrock_mantle/ap-southeast-2/google.gemma-3-12b-it, bedrock_mantle/ap-southeast-3/google.gemma-3-12b-it, bedrock_mantle/eu-central-1/google.gemma-3-12b-it, bedrock_mantle/eu-north-1/google.gemma-3-12b-it, bedrock_mantle/eu-south-1/google.gemma-3-12b-it, bedrock_mantle/eu-west-1/google.gemma-3-12b-it, bedrock_mantle/eu-west-2/google.gemma-3-12b-it, bedrock_mantle/google.gemma-3-12b-it, bedrock_mantle/sa-east-1/google.gemma-3-12b-it, crusoe/google/gemma-3-12b-it, databricks/databricks-gemma-3-12b, deepinfra/google/gemma-3-12b-it, gemma-3-12b, google-gemma-3-12b-it, google.gemma-3-12b-it, google/gemma-3-12b-it, google/gemma-3-12b-it:free, novita/google/gemma-3-12b-it, nvidia_nim/google/gemma-3-12b-it

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works