Gemma 4 26B A4B
VerifiedGoogle's open multimodal model for text, image, and video with 262k context.
About Gemma 4 26B A4B
The model combines a 26B parameter architecture with native support for image, text, and video modalities. Its 262144-token context window allows handling of long multimodal sequences in a single pass. As an open-weight release from Google, it provides direct access to weights for customization and local inference.
Typical usage includes multimodal content analysis, video understanding, and cross-modal generation tasks. Developers leverage its open nature for fine-tuning on domain-specific image-text-video datasets. The design emphasizes broad accessibility while maintaining strong performance on mixed-modality inputs.
Capabilities
How Gemma 4 26B A4B compares
Gemma 4 26B A4B (striped bar) vs other multimodal on intelligence, speed and price.
Price
USD per 1M output tokens · Lower is better · Gemma 4 26B A4B ranks #14 of 124
Sources: Artificial Analysis (intelligence, speed) · OpenRouter (price).
Best for
Extended Video Content Review
Processes hours of video input alongside transcripts to identify patterns, summarize events, and answer queries spanning the full duration.
Large-Scale Multimodal Reports
Integrates text, images, and charts from lengthy documents to generate reasoned summaries and extract cross-referenced insights.
Complex Instruction Execution
Follows detailed multi-step prompts that combine visual analysis with long-context text generation for tasks like research synthesis.
Strengths & limitations
Strengths
- +Large 256k-token context window
- +Native support for image, text, and video inputs
- +Efficient 26B-scale architecture
Limitations
- –No audio modality support
- –May trail larger models on complex reasoning tasks
- –Higher inference cost for video processing
Pricing by provider
Live per-provider pricing & uptime, routed via OpenRouter. Prices are USD per 1M tokens.
| Provider | Input /1M | Output /1M | Context | Uptime |
|---|---|---|---|---|
| DekaLLM(bf16) | $0.06 | $0.33 | 262K | 93.1% |
| DeepInfra(fp8) | $0.07 | $0.34 | 262K | 99.2% |
| Cloudflare | $0.10 | $0.30 | 256K | 99.8% |
| Ambient | $0.10 | $0.30 | 262K | — |
| SiliconFlow(fp8) | $0.12 | $0.40 | 262K | 99.9% |
| Parasail(bf16) | $0.13 | $0.40 | 262K | 99.4% |
| Novita(bf16) | $0.13 | $0.40 | 262K | 99.8% |
| NextBit(bf16) | $0.13 | $0.40 | 262K | 99.3% |
| $0.15 | $0.60 | 262K | 100.0% | |
| Venice(bf16) | $0.16 | $0.50 | 256K | 99.6% |
Cost calculator
Estimate what Gemma 4 26B A4B would cost for your usage.
Based on Gemma 4 26B A4B's $0.06/1M input · $0.33/1M output. Estimate only — actual cost varies by provider and caching.
Quick start
OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. Only the model slug changes between models.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1",
apiKey: process.env.OPENROUTER_API_KEY,
});
const completion = await client.chat.completions.create({
model: "google/gemma-4-26b-a4b-it",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(completion.choices[0].message.content);Model slug: google/gemma-4-26b-a4b-it
Editor's verdict
Gemma 4 26B A4B is Google's open-weight multimodal with a 262K-token context window.
At $0.33 per 1M output tokens, it is very cost-efficient for its class, served by 10 providers.
As an open-weight model you can self-host it or call it through a hosted API.
Best suited to large 256k-token context window and native support for image, text, and video inputs.
Frequently asked questions
The model supports a context length of 262144 tokens.
User reviews
Real, verified reviews from the community shape this model's rating.
Loading reviews…
Other Gemma models
Sibling versions in the Gemma family from Google.
Similar models
Other multimodal worth comparing.