Skip to main content

Gemma 4 26B A4B IT

Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at a fraction of the compute cost. Supports multimodal input including text, images, and video (up to 60s at 1fps). Features a 256K token context window, native function calling, configurable thinking/reasoning mode, and structured output support. Released under Apache 2.0.

Model card on Hugging Face
Input / output price
$0.00 / $0.00 per 1M
Context
262K tokens
Released
Apr 2, 2026
Agent setup

Paste into Claude Code, Cursor, or Codex

Help me set up the Respan gateway: fetch and follow https://www.respan.ai/docs/agent-setup.md. Use model "google_gemini_ai/gemma-4-26b-a4b-it".

Read docs

Providers

13 providers, cheapest first, OpenRouter last, priced per 1M tokens after gateway discounts.

ProviderInputOutput
Google Gemini AIgoogle_gemini_ai/gemma-4-26b-a4b-it
$0.00/M$0.00/M
EmpirioLabsempiriolabs/gemma-4-26b-a4b-it
$0.05/M$0.29/M
DeepInfradeepinfra/gemma-4-26b-a4b-it
$0.07/M$0.34/M
Featherlessfeatherless/gemma-4-26b-a4b-it
$0.07/M$0.34/M
Makoramakora/gemma-4-26b-a4b-it
$0.09/M$0.28/M

Is a provider down? OpenRouter status

Pricing

Averaged across 13 providers, in USD per 1M tokens after gateway discounts.

Average input
$0.09 per 1M
Average output
$0.32 per 1M
Average cache read
$0.08 per 1M

Cost by volume

Total cost from 0 to 1T tokens on the cheapest and most expensive provider.

Highest$0.15/MTeraLowest$0.00/MGoogle Gemini AI

Throughput and latency

30-day averages from requests through the Respan gateway. Under each figure, every model on the gateway by its average, this one in blue.

Output speed

Higher is better

111tokens/sacross providers

Faster than 80% of the other 599 models on the gateway

Time to first token

Lower is better

1.2sacross providers

Faster than 63% of the other 152 models on the gateway

Response latency

Lower is better

6.3sacross providers

Faster than 27% of the other 780 models on the gateway

Benchmarks

The Respan Index plus 12 published scores from 4 sources, each linked to where it was published.

18.5/ 100

Respan Index#141 of 176

176 ranked models. The shaded band is this model's 95% interval, 14.3 to 27.4. Qwen3 235B-A22B Thinking 2507 and Gemini 2.5 Flash fall inside it, so the gap is within the margin of error.
Gemma 4 26B A4B IT by category: Coding 4.1, Agents 13.3 (estimated), Reasoning 20.4, Knowledge & language 15.2, Human preference 50.0
  1. 141Gemma 4 26B A4B18.5
  2. 141Qwen3 235B-A22B Thinking 250718.5
  3. 141Gemini 2.5 Flash18.5

Published scores

BenchmarkScore
LiveCodeBench v6Google77.1%
WeirdMLHåvard Tveit Ihle35.2%

Compare Gemma 4 26B A4B IT: Gemma 4 26B A4B vs Gemma 4 31B, Gemma 4 26B A4B vs Gemini 4 Neon, Gemma 4 26B A4B vs Gemini 4 Carbon, Gemma 4 26B A4B vs Gemini 4 Hydrogen, Gemma 4 26B A4B vs Gemini 4 Argon, Gemma 4 26B A4B vs Gemini 3.8 Flash

Apps

Use Gemma 4 26B A4B IT in the tools you already run. Each guide points the app at the Respan gateway.

Setup guides for more apps are in the integration docs.

Gemma 4 26B A4B IT FAQ

Short answers, built from the data on this page.

Gemma 4 26B A4B IT is a model from Google Gemini AI, released on Apr 2, 2026. Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at a fraction of the compute cost. Supports multimodal input including text, images, and video (up to 60s at 1fps). Features a 256K token context window, native function calling, configurable thinking/reasoning mode, and structured output support. Released under Apache 2.0.

On the Respan gateway, Gemma 4 26B A4B IT costs from $0.00 per 1M input tokens and $0.00 per 1M output tokens on Google Gemini AI, after gateway discounts. Averaged across its 13 providers, it costs $0.09 per 1M input tokens and $0.32 per 1M output tokens.

Gemma 4 26B A4B IT is available through 13 providers: Google Gemini AI, EmpirioLabs, DeepInfra, Featherless, Makora, Nextbit256, and 7 more.

Gemma 4 26B A4B IT accepts up to 262K tokens of context on its largest provider route.

Point any OpenAI-compatible SDK at https://gateway.respan.ai/v1 with your Respan API key and set the model to "google_gemini_ai/gemma-4-26b-a4b-it". To set it up with a coding agent, copy the agent setup prompt at the top of this page.

Gemma 4 26B A4B IT scores 18.5 out of 100 on the Respan Index v1.1, ranking #141 of 176 models. By category: Coding 4.1, Agents 13.3 (estimated), Reasoning 20.4, Knowledge & language 15.2, and Human preference 50.0. The Respan Index combines independent and lab-reported benchmark results into one score, using benchmark data collected by Respan.

Gemma 4 26B A4B IT has 12 published benchmark scores, including LiveCodeBench v6 77.1%, GPQA Diamond 82.3%, and AIME 2026 (no tools) 88.3%. Each score links to where it was published; the data is collected by Respan.

Certified where it counts.

Built to meet the security and privacy standards that enterprise and healthcare teams require.

ISO 27001

The internationally recognized standard for information security management.

SOC 2

Secure, compliant management of your data across all of our systems.

GDPR

Operated under GDPR, the world's strictest standard for data privacy.

HIPAA

HIPAA compliant, with a BAA available for healthcare teams.

Works with your entire stack

Use Respan with your favorite frameworks and tools.

AnthropicAI21 LabsxAIVercel AI SDK
LangChainOpenRouterAssemblyAICursor
LiteLLMGroqBasetenMem0
CohereCogneeLlamaIndexFireworks
DeepSeekLinkupHaystackTogether AI
PerplexityInference.netPostHogRubyLLM
SlackMastraAzure OpenAIMistral
AWS BedrockMoonshotHyperspellBAML
ModaInstructorGoogle GeminiParasail
ReductoZapierAgnoNebius
OpenAINovita AIReplicateSuperagent

Built for AI agents.
Break less.
Ship more.