ZMIME
Google · Released Jul 13, 2024

Gemma 2 27B

Open Gemma instruction model for efficient chat and self-hosted deployments.

  • Open weights
Capability
122.1ECI · #152 of 165
Input
$0.65per 1M tokens
Output
$0.65per 1M tokens
Context
8K2K max output
Pricing

What it costs

Input
$0.65per million tokens
Output
$0.65per million tokens
Blended (3:1)
$0.65cheaper than 52% of priced models

Typical monthly bills

Monthly usageEstimated cost
Side project2M input + 0.5M output tokens$1.63
Team assistant25M input + 5M output tokens$19.50
Production app250M input + 50M output tokens$195.00

Median of 3 paid API providers on models.dev.

Capability

Independent benchmarks

122.1Epoch Capabilities Index
#152 of 165 scored models
88Median 146169

The shaded band is Epoch AI’s confidence range (115.8–124.1); the tick marks the median scored model.

  • GPQA DiamondGraduate-level science questions36.5%
  • OTIS Mock AIME 2024–2025Competition mathematics1.4%

Scores from Epoch AI, run independently of Google.

Specifications

The details

Lab
Google
Released
Jul 13, 2024
Knowledge cutoff
Jun 30, 2024
Context window
8,192 tokens
Max output
2,048 tokens
Inputs
Text
Output
Text
Reasoning
No
Tool calling
No
Structured output
No
Weights
Open weights
Availability
3 API providerslisted on models.dev
Same lab

More from Google

Questions

About Gemma 2 27B

How much does Gemma 2 27B cost?

Gemma 2 27B costs $0.65 input / $0.65 output per million tokens (median across 3 API providers). At a 3:1 input-to-output mix that is $0.65 per million tokens, cheaper than 52% of the 445 priced models we track.

What is the context window of Gemma 2 27B?

Gemma 2 27B accepts up to 8,192 tokens per request and can write up to 2,048 tokens in one response.

How good is Gemma 2 27B?

Epoch AI gives Gemma 2 27B a Capabilities Index score of 122.1 (likely range 115.8–124.1), ranking it #152 of 165 models Epoch has scored. Epoch AI benchmark results: GPQA Diamond 36.5%, OTIS Mock AIME 2024–2025 1.4%.

Is Gemma 2 27B open source?

Yes. Google publishes the weights, so it can be downloaded and self-hosted.

What inputs does Gemma 2 27B support?

Gemma 2 27B accepts text and replies in text. It is not a dedicated reasoning model, does not support tool calling.

When was Gemma 2 27B released?

Google released Gemma 2 27B on Jul 13, 2024. Its training data runs to Jun 30, 2024.

What are the best alternatives to Gemma 2 27B?

The closest current models from other labs on capability, price and release date are Mistral Large (Mistral AI, ECI 122.0, $0.50 / $1.50), Llama-3.1-70B-Instruct (Meta, ECI 125.9, $0.72 / $0.72), Claude Haiku 3 (Anthropic, ECI 118.8, $0.25 / $1.25) and GPT-4o mini (OpenAI, ECI 126.5, $0.15 / $0.60).