Gemma 4 26B A4B es un modelo multimodal abierto de Google con contexto 256K, texto, imagen y entrada de vídeo, herramientas y salida estructurada.
GoogleGeneracion de texto256K contextoLanzado Mar 31, 2026Inferencia nativaNuevoGemma 4 26B A4B es un modelo multimodal abierto de Google con contexto 256K, texto, imagen y entrada de vídeo, herramientas y salida estructurada.
Soporta texto, imagen y entrada de vídeo, streaming, herramientas de función, salida estructurada JSON, control de semillas y modo de pensamiento por defecto. Use reasoning effort o thinking presupuesto para el pensamiento ligado, o enable thinking=false para respuestas directas. Las lecturas automáticas de caché se facturan a la tasa de entrada de caché cuando se informa por el servicio modelo. No se admiten controles de caché explícitos.
También conocido como Google Gemma 4 26B-A4B, Gemma-4-26B-A4B
gemma-4-26b-a4bPOST /v1/chat/completionsPOST /v1/responsesPOST /v1/messagesPOST /v1/completionsLive pay-as-you-go rates from the EmpirioLabs catalog. You are billed only for what you use, with no monthly minimum.
Gemma 4 26B-A4B serves the OpenAI-compatible Chat Completions API. Point any OpenAI SDK at https://api.empiriolabs.ai/v1 with your EmpirioLabs API key and use the model id gemma-4-26b-a4b. Get an API key from the EmpirioLabs dashboard.
curl https://api.empiriolabs.ai/v1/chat/completions \
-H "Authorization: Bearer $EMPIRIOLABS_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gemma-4-26b-a4b",
"messages": [
{"role": "user", "content": "Write a haiku about the ocean."}
]
}'from openai import OpenAI
client = OpenAI(
base_url="https://api.empiriolabs.ai/v1",
api_key="YOUR_EMPIRIOLABS_API_KEY",
)
response = client.chat.completions.create(
model="gemma-4-26b-a4b",
messages=[{"role": "user", "content": "Write a haiku about the ocean."}],
)
print(response.choices[0].message.content)Request parameters supported by the Gemma 4 26B-A4B API on EmpirioLabs. Defaults apply when a field is omitted.
| Parámetro | Tipo | Predeterminado | Rango / valores | Descripción |
|---|---|---|---|---|
| temperature | number | 1 | 0 to 2 | Sampling temperature. Lower values are more deterministic. |
| top_p | number | 0.95 | 0 to 1 | Nucleus sampling probability mass. |
| max_tokens | integer | 4096 | 1 to 32768 | Maximum output tokens. |
| stop | string | - | - | One or more stop strings. |
| reasoning_effort | enum | medium | none, low, medium, high, max | Reasoning effort. none disables thinking; low, medium, high, and max set bounded thinking budgets. |
| enable_thinking | boolean | true | - | Enable the model reasoning channel before final output. |
| thinking_budget | integer | 4096 | 128 to 32768 | Maximum thinking tokens before the final answer. If max_tokens is lower, the service reserves room for the answer. |
| top_k | integer | 20 | 1 to 200 | Limit sampling to the top K candidate tokens when supported. |
| min_p | number | 0 | 0 to 1 | Minimum probability threshold for token sampling. |
| presence_penalty | number | 0 | -2 to 2 | Penalty for tokens that already appeared in the generated text. |
| frequency_penalty | number | 0 | -2 to 2 | Penalty based on how often a token has already appeared. |
| repetition_penalty | number | 1 | 0.1 to 2 | Penalty used by SGLang to reduce repeated text. |
| seed | integer | - | 0 to 2147483647 | Optional random seed for reproducible sampling. |
| logprobs | boolean | false | - | Return token log probabilities when supported. |
Soporta texto, imagen y entrada de vídeo, streaming, herramientas de función, salida estructurada JSON, control de semillas y modo de pensamiento por defecto. Use reasoning effort o thinking presupuesto para el pensamiento ligado, o enable thinking=false para respuestas directas. Las lecturas automáticas de caché se facturan a la tasa de entrada de caché cuando se informa por el servicio modelo. No se admiten controles de caché explícitos.
On EmpirioLabs, Gemma 4 26B-A4B is billed pay as you go: Input $0.05 (was $0.15) por señalización rápida 1M; Producto $0.29 (was $0.50) per 1M generados fichas; Caché implícita $0.025 (was $0.15) por fichas de entrada en caché de 1M. The live rate card on this page always matches what the API charges.
Gemma 4 26B-A4B supports a 256K-token context window with up to 32,768 output tokens per response.
Yes. Gemma 4 26B-A4B serves the OpenAI-compatible Chat Completions API, so existing OpenAI SDKs work by pointing base_url at https://api.empiriolabs.ai/v1 and setting the model id to gemma-4-26b-a4b.
Yes. The EmpirioLabs playground runs Gemma 4 26B-A4B in the browser with the same parameters the API exposes, so you can test prompts before writing code.
Create an EmpirioLabs account, then generate a key under API Keys in the dashboard. Billing is pay-as-you-go credits, so you only pay for the requests you make.
Check out our pricing or reach out if you want your own model deployed on our stack.
