Gemma 4 26B A4B
Provider: Google AI US Official API price
Input price
$0.15
/ 1M tokens
Output price
$0.60
/ 1M tokens
Cache price
$0.037
/ 1M tokens (read)
Context
262K tokens
Max output 236K tokens
Overview
Google DeepMind's Gemma 4 26B A4B is a mid-tier instruction-tuned MoE model built for general multimodal reasoning and function calling.
Strengths
- 25.2B total parameters with 3.8B active per token
- 262,144 token context window and 235,929 max output
- Native support for vision, JSON mode, and thinking
- Available for free via Google AI Studio and OpenRouter
Best use cases
- Processing large documents within the 262K context
- Structured data extraction using JSON mode and functions
- Multimodal analysis leveraging built-in vision support
Pricing position
Input is $0.15/1M tokens and output is $0.60/1M tokens, matching GPT-4o-mini and Mistral Small 4 while doubling the context window of both.
Caveats
Free-tier prompts on Google AI Studio may be used for model training.
Notes drafted by an LLM from the figures on this page (2026-11-10) and rewritten when the price changes.
Key facts
- Price basis
- Official API price
- Max output
- 236K tokens
- Listed on OpenRouter
- 2026-04-03
- Tracked since
- 2026-05-08
- Last updated
- 2026-11-10
Features
Image inputFunction callingJSON outputReasoning
Free tier
| Free via | Limits | Note |
|---|---|---|
| Google AI Studio Maker’s API | Not published (shown in the provider console) | Prompts may be used to improve the provider’s products |
| OpenRouter (:free) Third-party host | RPM 20 · RPD 50 | 1,000/day after buying $10 of credits |
Monthly cost estimate
| Scale | Input (M) | Output (M) | Monthly |
|---|---|---|---|
| 0.1M (hobby) | 0.1 | 0.05 | $0.04 |
| 1M (small app) | 1 | 0.5 | $0.45 |
| 10M (mid scale) | 10 | 5 | $4.50 |
| 100M (large) | 100 | 50 | $45.00 |
※ Cache and batch discounts not factored in. Detailed calculator
Official resources
Other models from Google AI
Similar-priced models
Source: OpenRouter API (google/gemma-4-26b-a4b-it)