Gemma 4 26B A4B
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...
Cost / 1M tokens
Very low usage$0.076
input
$0.043
cached input
$0.255
output
Context Window
262K
Up to 236K max output
Modalities
image, text, video → text
Gemma tokenizer
Quickstart & API
google/gemma-4-26b-a4b-itBest for & Strengths
Key Strengths
Recommended Workloads in ChatOSS
- Coding agents
- Complex reasoning & debugging
- Screenshot-to-code workflows
Verified Evaluations
Benchmarks & Frontier Comparison
How Gemma 4 26B A4B compares against leading open-weight and proprietary frontier models across independent agentic coding and reasoning suites.
| Benchmark (Metric) | Gemma 4 26B A4B | DeepSeek V4.1 Flash | GLM 5.3 | Kimi K3.1 Pro | Claude Opus 5.5 |
|---|---|---|---|---|---|
| Agentic & Software Engineering | |||||
| ChatOSS Agentic Coding Index (%) | — | 63.2% | 58.3% | 65% | 74.1% |
| SWE-Bench Verified (%) | — | 91.5% | 95.4% | 94.8% | 97% |
| DeepSWE v1.1 (Pass@1 %) | — | 77% | 69% | 73.5% | 74.2% |
| Vibe Code Bench v1.1 (%) | — | 87.7% | 78.1% | 88.4% | 90.3% |
| Terminal-Bench 4.0 (%) | — | 42.3% | 42% | 46.2% | 65.2% |
| Code Migration (%) | — | 45.6% | 44.2% | 52% | 66.7% |
| Intelligence & Token Economics | |||||
| Artificial Analysis Intelligence Index (0–100) | 16.7 | 39.5 | 44.8 | 55.1 | 57.6 |
| Input Price (/ 1M tokens) | $0.076 | $0.15 | $1.40 | $2.00 | $4.00 |
| Output Price (/ 1M tokens) | $0.255 | $0.60 | $4.40 | $10.00 | $20.00 |
Artificial Analysis Intelligence Index
Artificial Analysis
Active Provider Endpoints (15)
Upstream inference providers serving Gemma 4 26B A4B , with per-provider pricing, context limits, quantization, and 24-hour uptime.
| Provider | Input / 1M | Output / 1M | Cache Read / 1M | Context | Max Output | Quantization | Uptime (24h) |
|---|---|---|---|---|---|---|---|
Darkbloomcache darkbloom | $0.042 | $0.22 | $0.021 | 131K | 33K | — | 100.0% |
DekaLLMcache dekallm/bf16 | $0.06 | $0.33 | $0.04 | 262K | 236K | bf16 | 98.5% |
AkashMLcache akashml/fp4 | $0.07 | $0.24 | $0.046 | 131K | 118K | fp4 | 95.7% |
Makoracache makora | $0.07 | $0.24 | $0.046 | 256K | 128K | — | 99.9% |
DeepInfra deepinfra/fp8 | $0.07 | $0.34 | — | 262K | 16K | fp8 | 99.9% |
NextBitcache nextbit/bf16 | $0.076 | $0.255 | $0.043 | 262K | 236K | bf16 | 99.7% |
Io Netcache io-net/bf16 | $0.09 | $0.31 | $0.041 | 262K | 236K | bf16 | 100.0% |
PrimeIntellect primeintellect/bf16 | $0.10 | $0.30 | — | 262K | 33K | bf16 | 99.9% |
CoreWeavecache coreweave/bf16 | $0.10 | $0.30 | $0.05 | 262K | 236K | bf16 | 100.0% |
Cloudflarecache cloudflare | $0.10 | $0.30 | $0.05 | 256K | 230K | — | 100.0% |
Venicecache venice/bf16 | $0.13 | $0.40 | $0.05 | 256K | 8K | bf16 | 100.0% |
Parasailcache parasail/bf16 | $0.13 | $0.40 | $0.05 | 262K | 236K | bf16 | 100.0% |
Novita novita/bf16 | $0.13 | $0.40 | — | 262K | 131K | bf16 | 100.0% |
SiliconFlowcache siliconflow/fp8 | $0.14 | $0.40 | $0.05 | 262K | 236K | fp8 | 94.0% |
Google google-vertex/global | $0.15 | $0.60 | — | 262K | 236K | — | 93.9% |
Supported API Parameters
Parameters accepted by this model when called through ChatOSS or OpenRouter.
Compare Similar Models
Browse all models →Gemini 4 Argon
Gemini 4 Argon is Google's flagship reasoning model, built for complex software engineering, deep mathematical and scientific problem solving, and autonomous multi-step agents. Google trained it to keep context across...
Gemini 3.8 Flash
Gemini 3.8 Flash is Google's most intelligent Flash model with significant gains from 3.7 Flash across software engineering, agentic tasks, and multi-step reasoning.
Gemini 3.7 Flash
Gemini 3.7 Flash is a multimodal model from Google for fast agentic workflows, coding, and complex multi-step reasoning. It is designed for tasks that require responsive performance and reliable multi-step...
Gemini 3.6 Flash
Gemini 3.6 Flash is a high-efficiency model from Google for coding, agentic workflows, and web and app development. It is designed to produce polished outputs with fewer unnecessary edits and...
Anthropic
Claude Opus 5.5
Claude Opus 5.5 is Anthropic's flagship model for demanding reasoning, coding, and long-horizon agentic work, succeeding Claude Opus 5. It is particularly strong at multi-step changes in large codebases, code...
Moonshot AI
Kimi K3.1 Pro
Kimi K3.1 Pro is Moonshot AI's flagship open-weight multimodal reasoning model and a post-trained successor to Kimi K3, with a 1M-token context window. It is built for long-horizon coding, 3D...