GoogleReleased on Mar 12, 2025

Gemma 3 27B: API Pricing, Context Window & Benchmarks

Gemma 3 27B is a language model from Google, released in March 2025, with multimodal input.

Gemma 3 27B is a 27-billion-parameter vision-language model from Google, handling text and image input and generating text output. It features a 128K context window, multilingual support, and open weights. Suitable for complex question

Input
TextImage
Output
Text

Gemma 3 27B benchmarks

Rankings

Quality Tracker

Gemma 3 27B Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

LLM Stats Logollm-stats.com - Tue Jul 28 2026
Notice missing or incorrect data?

Gemma 3 27B pricing

Providers

Gemma 3 27B starts at $0.100 per million input tokens and $0.200 per million output tokens via DeepInfra. See all 2 providers below with their per-token pricing, latency, throughput, and modality support.

ProviderInput $/MOutput $/MContext in / outTTFT p50 / p95 sOutput avg / p5 c/sSuccess 7dModalities in / out
DeepInfra logoDeepInfra
$0.100$0.200131.1K/131.1K
/0.20
33/
/
Novita logoNovita
$0.110$0.200131.1K/131.1K
/0.20
33/
/

Cached input is the discounted price for prompt tokens served from a provider cache. TTFT is time to first token. Output is characters per second; p5 is the sustained floor exceeded by 95% of observed requests. Success is calculated from completed versus failed requests over the trailing seven days.

Loading chart...
Loading chart...
Loading chart...

Gemma 3 27B model size

Gemma 3 27B has 27 billion parameters and was trained on 14 trillion tokens. See how it compares to other models in the same parameter range.

ParametersTraining tokens
27B
14Ttokens
519× tokens-to-params ratio
Medium (10–30B)
27B
1B7B70B405B

Gemma 3 27B context window

Input and output token limits for Gemma 3 27B, plus how it ranks on long-context understanding.

InputOutput
131Ktokens
131Ktokens
197 pages of text
131K
8K128K1M

Gemma 3 27B latency

Gemma 3 27B time to first token, sustained output throughput, and failed-request rate from live API traffic over the trailing 7 days.

Gemma 3 27B examples

Recent arena outputs from Gemma 3 27B, picked from the highest-ranked matchups.

Gemma 3 27B license

Gemma 3 27B is released under the Gemma license, which permits commercial use, has 27.0B parameters.

License
Gemma
Commercial use allowed
Parameters
27.0B

Google Gemma Terms of Use

Gemma 3 27B resources

Official sources for Gemma 3 27B: paper or system card, official launch post, model weights.

Gemma 3 27B vs other models

The most-compared alternatives to Gemma 3 27B are Kimi K2 0905, MiMo-V2.5-Pro, Llama 3.1 70B Instruct. Open any pair side-by-side for benchmarks, pricing, context, and latency.

Models like Gemma 3 27B

Models ranked just above and below Gemma 3 27B by LLM Stats score.

 

Kimi K2 0905

Score pending
 

MiMo-V2.5-Pro

Score pending
 

Llama 3.1 70B Instruct

Score pending
 

Nova Lite

Score pending
 

Gemini 1.5 Flash

Score pending
 

Gemini 2.0 Flash

Score pending

FAQ

Common questions about Gemma 3 27B.

When was Gemma 3 27B released?

Gemma 3 27B was released on March 12, 2025 by Google. This is the official Gemma 3 27B release date tracked on LLM Stats.

How much does Gemma 3 27B cost?

Gemma 3 27B pricing starts at $0.10 per million input tokens and $0.20 per million output tokens via DeepInfra, the lowest price among tracked providers.

How big is Gemma 3 27B?

Gemma 3 27B has 27 billion parameters. It was trained on 14.0 trillion tokens. It ships as an open-weight model, so you can download and run it on your own hardware.

Who created Gemma 3 27B?

Gemma 3 27B was created by Google.

What is the license for Gemma 3 27B?

Gemma 3 27B is released under the Gemma license. This is an open-source / open-weight license that permits self-hosting.

Is Gemma 3 27B multimodal?

Yes, Gemma 3 27B is multimodal and can accept both text and images as input.

What is Gemma 3 27B latency?

Gemma 3 27B p95 time to first token is 0.20 seconds via DeepInfra over the trailing 7 days. Lower time to first token means the model begins responding sooner for chat, agents and API workloads.

Where can I use Gemma 3 27B?

Gemma 3 27B is available through 2 providers including DeepInfra, Novita.

Where is the Gemma 3 27B paper or technical report?

Gemma 3 27B has a paper or technical report available at https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf. Use that source for architecture, training, release and evaluation details.

What models should I compare Gemma 3 27B against?

Common Gemma 3 27B comparisons include Gemma 3 27B vs Kimi K2 0905, Gemma 3 27B vs MiMo-V2.5-Pro, Gemma 3 27B vs Llama 3.1 70B Instruct. Compare them side by side for benchmark scores, pricing, context window, latency and API availability.