TheRouter.aiTheRouter.ai
Back to Models

Gemma 3 12B

googlegoogle/gemma-3-12b

Google's open-source Gemma 3 12B model with vision. Efficient and fast for everyday tasks.

Gemma 3 model card ↗Introducing Gemma 3 ↗Gemma 3 weights on Hugging Face ↗

Gemma 3 12B is the mid-sized member of Google's Gemma 3 open-model family, announced on March 12, 2025. It is a 12-billion-parameter vision-language model with image and text input, text output, a 128K-token context window, and the Gemma community's standard instruction-tuned variant for chat-style use. Google positions Gemma 3 as its most capable open family at launch, with longer context, multimodality, function calling, structured outputs, and coverage for more than 140 languages.

For TheRouter, Gemma 3 12B is a pragmatic open-weight choice when a team wants image understanding and multilingual chat at lower cost than frontier hosted models. It is stronger than small edge models for document QA, visual inspection, extraction, and general assistant workloads, while still being small enough for local or dedicated deployments. It should not be treated as a Gemini replacement: the best fit is controlled mid-complexity work where predictable cost and open-weight portability matter more than peak reasoning.

Best for
  • • Vision-language extraction, inspection, and image-grounded support flows that need a 12B open-weight model rather than a proprietary frontier model
  • • Multilingual assistants and content workflows where 140+ language coverage matters and model portability is a buying criterion
  • • Cost-sensitive document QA and extraction with long but bounded context, especially when 128K tokens is enough
  • • Teams evaluating open-weight deployment paths before deciding whether to self-host, fine-tune, or keep a managed API route
Reach for something else if
  • • Frontier reasoning, autonomous coding, or hard math where Gemini 2.5 Pro, Gemini 3-class models, Claude, or GPT-5.5-class models have a much higher ceiling
  • • Ultra-cheap routing, classification, or extraction where smaller Gemma, Llama, or Qwen models can meet quality targets at lower latency
  • • Native image, audio, or video generation — Gemma 3 12B understands images but returns text, so use a dedicated generation model

How TheRouter serves this differently from the vendor

As the vendor operates it

Google publishes Gemma 3 12B as open weights for self-hosting, fine-tuning, and deployment across local hardware, Hugging Face, Kaggle, Google AI Studio, Vertex AI, and other runtimes.

On TheRouter

TheRouter serves it as a managed OpenAI-compatible chat-completions route with text/image input and text output; this page does not claim local weight download, fine-tuning, or vendor-native runtime controls on the live TheRouter request path.

Google Developers Blog ↗
Context Length
128K
Max Output
8K
Input Priceper 1M tokens
$0.0972/ 1M tokens
Output Priceper 1M tokens
$0.3132/ 1M tokens

Modalities

textimage→text

Capabilities

Vision

Pricing Breakdown

TypeRate
Input$0.0972 / 1M tokens
Output$0.3132 / 1M tokens

Supported Parameters

temperaturemax_tokenstop_pstop

Specifications

Release date2025-03-12Google Developers Blog ↗verified
Architecture / sizeDense transformer, 12B parametersGoogle Developers Blog ↗verified
Context windowUp to 128K tokensGoogle Developers Blog ↗verified
Training tokens12T tokens for the 12B modelGoogle Developers Blog ↗verified
Input / output modalitiesText and images in; text outGoogle Developers Blog ↗verified
LicenseGemma Terms of Useai.google.dev ↗verified

Benchmarks

BenchmarkDistributionScoreSource
MMLU Pro
Gemma 3 IT 12B, 0-shot, as reported in Google's Gemma 3 model card
60.6%ai.google.dev ↗
HumanEval
Gemma 3 IT 12B, 0-shot, code generation benchmark from the official model card
85.4%ai.google.dev ↗
GSM8K
Gemma 3 IT 12B, 0-shot, grade-school math benchmark from the official model card
94.4%ai.google.dev ↗

API Usage Examples

Use the global api.therouter.ai endpoint shown below for new integrations; the legacy China accelerated endpoint is retired.

cURL
curl https://api.therouter.ai/v1/chat/completions   -H "Content-Type: application/json"   -H "Authorization: Bearer $THE_ROUTER_API_KEY"   -d '{
    "model": "google/gemma-3-12b",
    "messages": [
      {"role": "user", "content": "Summarize the key points from this input."}
    ]
  }'

API guide

Full API reference →

Chat completion

cURL
curl https://api.therouter.ai/v1/chat/completions \
  -H "Authorization: Bearer $THEROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemma-3-12b",
    "messages": [{"role": "user", "content": "Explain adaptive image tiling in two paragraphs."}]
  }'

More from google

google/gemini-2.5-flashManaged Google workhorse with stronger reasoning, longer context, and native Gemini toolinggoogle/gemini-2.5-proGoogle's higher-ceiling reasoning option when quality matters more than open weightsgoogle/gemini-3.5-flashNewer Gemini Flash route for teams prioritizing managed frontier throughput

Similar models

Cross-provider sibling models
mistral/ministral-14b
Similar open-weight 14B-class vision-language competitor with Apache 2.0 licensing
meta/llama-3.1-8b
Smaller open model for lower-latency text workloads without Gemma 3's vision focus
qwen/qwen3-14b
Alibaba's 14B-class open model alternative with strong multilingual and reasoning coverage

News & changes

2025-03-12

Google introduces Gemma 3 with multimodality and 128K context

Google's launch post presents Gemma 3 as a four-size open-model family, with 1B, 4B, 12B, and 27B variants, image-and-text input, text output, structured outputs, function calling, and more than 140 supported languages.

re-authored by TheRouterGoogle Developers Blog ↗

Frequently asked

Is Gemma 3 12B a good default model for production chat?

Use it when open weights, visual input, and predictable cost matter. For highest reasoning quality, longer context, or managed Google-native tools, choose a Gemini route instead.

Does Gemma 3 12B generate images?

No. Gemma 3 12B is a vision-language model that can read image input and return text. Use a dedicated image model for generation.

How is TheRouter serving different from Google's open-weight release?

Google's release lets teams download and run weights in multiple environments. TheRouter exposes a managed OpenAI-compatible API route, so integration is simpler but local runtime and fine-tuning controls are outside this route.

Are the API snippets verified?

No. This curator run has no paid API budget, so snippet verification remains unproven until a human operator approves and funds the verification run.

Fact ledger — every claim on this page traces here
sourceURLretrieved
Release dateGoogle Developers Blog ↗2026-08-21verified
Architecture / sizeGoogle Developers Blog ↗2026-08-21verified
Context windowGoogle Developers Blog ↗2026-08-21verified
Training tokensGoogle Developers Blog ↗2026-08-21verified
Input / output modalitiesGoogle Developers Blog ↗2026-08-21verified
Licenseai.google.dev ↗2026-08-21verified
MMLU Proai.google.dev ↗2026-08-21verified
HumanEvalai.google.dev ↗2026-08-21verified
GSM8Kai.google.dev ↗2026-08-21verified
Google introduces Gemma 3 with multimodality and 128K contextGoogle Developers Blog ↗2026-08-21verified
Help & contact