impulse

Gemma 3 12B

Open weights
API id databricks-gemma-3-12b
Input / 1M
$0.15
at Databricks
Output / 1M
$0.50
at Databricks
Context window
128K
tokens
Max output
32K
tokens per reply
Hosts
1
1 listings
Released
-
date not published

Pricing

List prices at Databricks, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.15$0.50
Cached input$0.15
Cache write, 5 minutes$0.15

What it costs a month

At Databricks list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$82.50
  • RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$90.01
  • Coding agent1,000 requests a day, 40K in / 2K out, cached$210
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$480
Use your own numbers

Where to run it

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

HostIn / out per 1M
databricks-gemma-3-12b
$0.15 / $0.50

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
Text
Produces
Text
Open weights
License: Gemma License
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 Databricks.

Smaller Google models

No cheaper Google model keeps its key capabilities.

Other frontier labs

  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    16% less
  • GPT-6 Luna
    OpenAI · on OpenAI
    $0.10 / $0.50
    Has every key capability
    16% less

Open-weight picks

  • Qwen3.8 Omni Flash
    Alibaba Qwen · on Alibaba DashScope
    $0.15 / $0.47
    Has every key capability
    3% less
  • Ministral 14B 2512
    Mistral AI · on Mistral AI
    $0.20 / $0.20
    Has every key capability
    16% less
  • Nemotron 3.5 Super VL
    NVIDIA · on Kyma API
    $0.081 / $0.472
    Has every key capability
    25% less
  • Gemma 4 31B IT
    Google · on W&B Inference
    $0.10 / $0.34
    Has every key capability
    33% less

Call it

Using Databricks's API.

Python
from litellm import completion

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

Quick answers

How much does Gemma 3 12B cost?

On Databricks, Gemma 3 12B costs $0.15 per million input tokens and $0.50 per million output tokens. Cached input costs $0.15 per million tokens.

Which APIs offer Gemma 3 12B?

Gemma 3 12B runs on 1 API: Databricks.

What is Gemma 3 12B's context window?

128K tokens, with up to 32K tokens in one reply.

Is Gemma 3 12B being retired?

No retirement date is scheduled for it on any API we track.

What are cheaper alternatives to Gemma 3 12B?

Models that cost less and keep most of what it can do: Claude Haiku 5.5 on Anthropic (16% less), GPT-6 Luna on OpenAI (16% less) and Qwen3.8 Omni Flash on Alibaba DashScope (3% less). Savings are on a mix of 3 input tokens to 1 output token.

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

Use databricks-gemma-3-12b with Databricks.

Sources

Raw catalog entry for databricks-gemma-3-12b on Databricks
{
  "cache_creation_input_token_cost": 1.5001e-7,
  "cache_read_input_token_cost": 1.5001e-7,
  "input_cost_per_token": 1.5001e-7,
  "input_dbu_cost_per_token": 0.0000021429999999999996,
  "litellm_provider": "databricks",
  "max_input_tokens": 128000,
  "max_output_tokens": 32000,
  "metadata": {
    "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."
  },
  "mode": "chat",
  "output_cost_per_token": 5.0001e-7,
  "output_dbu_cost_per_token": 0.000007143,
  "source": "https://www.databricks.com/product/pricing/foundation-model-serving"
}