impulse

Gemma 3 12B IT

Open weights

Gemma 3 introduces multimodality, supporting vision-language input and text outputs.

API id gemma-3-12b-itReleased 13 Mar 2025 (Mar 2025)
A newer model in this line is out: Gemma 4 12B IT.
Amazon Bedrock retires it on 30 Mar 2027 (in 132 days). 4 hosts keep serving it.
Input / 1M
$0.05
at Novita AI
Output / 1M
$0.10
at Novita AI
Context window
131K
tokens
Max output
8K
tokens per reply
Hosts
5
6 listings
Released
Mar 2025
13 Mar 2025

Pricing

List prices at Novita AI, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.05$0.10

What it costs a month

At Novita AI list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out$22.50
  • RAG over documents2,000 requests a day, 8K in / 0.6K out$27.60
  • Coding agent1,000 requests a day, 40K in / 2K out$66.00
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$144
Use your own numbers

Where to run it

5 hosts serve Gemma 3 12B IT. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
Novita AIInference
gemma-3-12b-it
$0.05 / $0.10
DeepInfraInference
gemma-3-12b-it
$0.05 / $0.15
+20% vs headline
CrusoeInference
gemma-3-12b-it
$0.10 / $0.10
+60% vs headline
google.gemma-3-12b-it
$0.09 / $0.29
+124% vs headline
Retires 30 Mar 2027
NVIDIA NIMInference
gemma-3-12b-it
-
Show 1 more listings: regions, snapshots and variants
Amazon Bedrock
Mantle
google.gemma-3-12b-it
$0.09 / $0.29
+124% vs headline
Retires 30 Mar 2027

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
TextImages
Produces
Text
Open weights
License: Gemma License
Runs in about 29.2 GB of GPU memory
Runtimes: Ollama, llama.cpp, vLLM

Cheaper alternatives

Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.

Same model, cheaper host

No host sells it for less than Novita AI.

Smaller Google models

No cheaper Google model keeps its key capabilities.

Other frontier labs

No cheaper match from another frontier lab.

Open-weight picks

  • Qwen3.7 Flash Variant1
    Alibaba Qwen · on EmpirioLabs AI
    $0.028 / $0.11
    Has every key capability
    22% less
  • Gemma3 4B
    Google · on LlamaGate
    $0.03 / $0.08
    Has every key capability
    32% less
  • Ministral 3B
    Mistral AI · on Azure AI Foundry
    $0.04 / $0.04
    Has every key capability
    36% less
  • Llama 3.2 3B
    Meta · on LlamaGate
    $0.04 / $0.08
    Lacks vision
    20% less

Model line

Newer and older versions, at each lab's own price.

  1. UndatedGemma 4 12B ITLatest$0.30 / $2.00
  2. Mar 2025Gemma 3 12B ITThis page$0.05 / $0.10

Call it

Using Novita AI's API.

Python
from litellm import completion

response = completion(
    model="novita/gemma-3-12b-it",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Gemma 3 12B IT cost?

On Novita AI, Gemma 3 12B IT costs $0.05 per million input tokens and $0.10 per million output tokens.

Which APIs offer Gemma 3 12B IT?

Gemma 3 12B IT runs on 5 APIs: Novita AI, DeepInfra, Crusoe, Amazon Bedrock and NVIDIA NIM.

What is Gemma 3 12B IT's context window?

131K tokens, with up to 8K tokens in one reply.

Is Gemma 3 12B IT being retired?

Amazon Bedrock retires it on 30 Mar 2027. 4 hosts keep serving it.

Is there a newer version of Gemma 3 12B IT?

Yes. Gemma 4 12B IT is the newest model in this line.

What are cheaper alternatives to Gemma 3 12B IT?

Models that cost less and keep most of what it can do: Qwen3.7 Flash Variant1 on EmpirioLabs AI (22% less). Savings are on a mix of 3 input tokens to 1 output token.

When was Gemma 3 12B IT released?

Google released it on 13 Mar 2025.

What is Gemma 3 12B IT's API model id?

Use gemma-3-12b-it with Novita AI. Other hosts use their own ids: google.gemma-3-12b-it on Amazon Bedrock.

Sources

Raw catalog entry for gemma-3-12b-it on Novita AI
{
  "litellm_provider": "novita",
  "mode": "chat",
  "input_cost_per_token": 5e-8,
  "output_cost_per_token": 1e-7,
  "max_input_tokens": 131072,
  "max_output_tokens": 8192,
  "supports_tool_choice": true,
  "supports_vision": true,
  "supports_system_messages": true,
  "supports_response_schema": true
}