Gemma 4 26B A4B API Benchmarks, Pricing & Provider Data
Compare Gemma 4 26B A4B with another model
Choose a model to open its comparison page.
Gemma 4 26B A4B API pricing covers undefined API provider} other undefined API providers}}, from $0.00005/request to $10.27/M.
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 qual...
Specifications
Input and output token limits for this model, plus how it ranks on long-context understanding.
Features
Technical Details
- Input
- Output
- Total parameters
- 26B
- Active parameters
- 4B
- Released
- Apr 2026
- Tokenizer
- Gemma
- Architecture
- text+image+video->text
- Moderated
- No
- Supported parameters
- frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p
OpenRouter endpoints
11 endpointsThird-party OpenRouter endpoint data, shown separately from LMSpeed measurements. Some 30-minute live performance fields only appear after syncing with an OpenRouter API key.
| Provider endpoint | Input | Output | 1d uptime | 30m latency | 30m throughput | Context / output |
|---|---|---|---|---|---|---|
Darkbloom darkbloom | $0.042/M | $0.220/M | 99.9% | — | — | undefined tokens / undefined tokens |
NextBit nextbit/bf16 | $0.090/M | $0.300/M | 99.8% | — | — | undefined tokens / undefined tokens |
Novita novita/bf16 | $0.130/M | $0.400/M | 99.7% | — | — | undefined tokens / undefined tokens |
Cloudflare cloudflare | $0.100/M | $0.300/M | 99.5% | — | — | undefined tokens / undefined tokens |
DekaLLM dekallm/bf16 | $0.060/M | $0.330/M | 99.5% | — | — | undefined tokens / undefined tokens |
DeepInfra deepinfra/fp8 | $0.070/M | $0.340/M | 99.5% | — | — | undefined tokens / undefined tokens |
Parasail parasail/bf16 | $0.130/M | $0.400/M | 99.0% | — | — | undefined tokens / undefined tokens |
SiliconFlow siliconflow/fp8 | $0.140/M | $0.400/M | 98.3% | — | — | undefined tokens / undefined tokens |
Venice venice/bf16 | $0.130/M | $0.400/M | 98.0% | — | — | undefined tokens / undefined tokens |
Makora makora | $0.100/M | $0.340/M | 97.5% | — | — | undefined tokens / undefined tokens |
Google google-vertex/global | $0.150/M | $0.600/M | 97.5% | — | — | undefined tokens / undefined tokens |
Pricing Comparison
Compare Gemma 4 26B A4B API pricing across 8 providers. Prices range from $0.00005/request to $10.27/M. CM-API 公益站 offers the lowest rate at $0.00005/request.
| Provider | Health | Model Variant | Group | Input ($/M) | Output ($/M) | Speed (t/s) | First token | Audit |
|---|---|---|---|---|---|---|---|---|
L1 100% | gemma-4-26b-a4b-it | Translate | $0.021/M | $0.064/M | — | — | — | |
L1 99% | google/gemma-4-26b-a4b-it:free | free | $0.00005/request | - | — | — | — | |
L1 99% | gemma-4-26b-a4b-it | default | $0.250/M | $0.500/M | — | — | — | |
L1 99% | gemma-4-26b-a4b-it | default | $0.137/request | - | — | — | — | |
L1 100% | google/gemma-4-26b-a4b-it | openrouter | $0.012/M Cache read$0.0023/M | $0.060/M | — | — | — | |
L1 0% | google/gemma-4-26b-a4b-it:free | default | $0.010/request | - | — | — | — | |
L1 0% | google/gemma-4-26b-a4b-it:free | OpenAI | $10.27/M | $10.27/M | — | — | — |
Alternatives & Similar Models
Gemma 4 31B
gemma-4-31b-it
Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function...
GPT-5.4
gpt-5-4
OpenAI GPT-5.4 extends the GPT-5 family with stronger instruction following, deeper tool use, and improved performance on coding, math, and long-document analysis.
GPT-5.2
gpt-5-2
OpenAI GPT-5.2 is a GPT-5 series model emphasizing advanced reasoning, multimodal understanding, and high-quality outputs for complex enterprise workloads.
Kimi K2.5
kimi-k2-5
Moonshot Kimi K2.5 is an open-weight multimodal agent model with native vision and text input, strong coding performance, and a 256K context window.
DeepSeek V3.2
deepseek-v3-2
DeepSeek V3.2 is an upgraded V3-series MoE model with stronger reasoning, coding, and math performance, widely available through OpenAI-compatible API relays.
MiniMax M2.7
minimax-m2-7
MiniMax M2.7 is a high-tier M2-series model tuned for complex reasoning, long-context dialogue, and production-grade API workloads.
Frequently Asked Questions
- What benchmark data does Gemma 4 26B A4B include?
- LMSpeed shows Gemma 4 26B A4B benchmark context, API price, output speed, first-token latency, and provider data across 8 providers when those signals are available.
- What is the Gemma 4 26B A4B API price?
- Gemma 4 26B A4B has pricing from undefined provider} other undefined providers}}, ranging from $0.00005/request to $10.27/M. CM-API 公益站 has the lowest listed price.
- What does the Gemma 4 26B A4B API pricing table include?
- The Gemma 4 26B A4B API pricing table compares 8 providers by input price, output price, free tier status, speed, first-token latency, and recent health data when available.
- Which provider has the cheapest Gemma 4 26B A4B API pricing?
- CM-API 公益站 currently has the lowest listed Gemma 4 26B A4B price at $0.00005/request across undefined provider} other undefined providers}}.
- Is Gemma 4 26B A4B API free?
- Gemma 4 26B A4B does not currently have a free API tier on LMSpeed. All 8 providers charge per token.