Skip to content
← Models
Open weightsIBM/10 trustMetrics-checked

Granite 4.0 H Tiny

Compact hybrid Mamba-2/Transformer model for efficient inference, Apache 2.0.

Visit official page →

About

Granite 4.0 H Tiny is the small sibling of IBM's Granite 4.0 hybrid family, released October 2025 alongside the Small, Micro, and Nano variants. It carries 7B total parameters with 1B active per token, using the same Mamba-2 state-space plus transformer hybrid architecture as the larger H models. IBM positions it for high-volume, low-complexity enterprise jobs where inference cost dominates.

Like the rest of the Granite 4.0 generation it is released under Apache 2.0 with signed checkpoints, and IBM reports the hybrid design delivers large memory savings on long inputs and concurrent batches. It runs natively on vLLM, llama.cpp, and MLX, making it a practical open option for cost-sensitive deployments.

Specs

Context window

Pricing

Free quota

API access

Weights

Open weights

Released

October 2, 2025

Benchmarks

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

high-volume-inferenceedge-deploymentchatbotssummarization

Trust score — 7/10

/10 trustMetrics-checked

Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 133; Trending score 74/100 from researcher assessment.

Sources