DiffusionGemma 26B
Last verified: June 11, 2026< $0.0001/1M input · < $0.0001/1M output · 262K context · Google
Count Tokens
—
Tokens
—
Input Cost
—
Output Cost
Estimate Monthly Cost
Monthly Cost Estimator
Quick:
< $0.0001/mo
Pick a preset above or enter custom usage
Alternatives to DiffusionGemma 26B
Pricing Details
GoogleOpen-weight (Apache 2.0). Experimental text-diffusion model on the Gemma 4 26B A4B backbone: 25.2B total, 3.8B active (MoE). Denoises a 256-token canvas in parallel for up to 4x faster generation (1,000+ tok/s on H100). Text/image/video in, text out; no audio. NVFP4 fits in 18GB VRAM. No per-token API; self-host (vLLM, SGLang, MLX) or Vertex AI. Scores below autoregressive Gemma 4 on every published benchmark.
Input / 1M tokens
$0
Output / 1M tokens
$0
Context Window
262K
Max Output
33K
Price History
Launched at current price on 2026-06-10. No price changes recorded.
Frequently Asked Questions
Common questions about DiffusionGemma 26B pricing and usage