AI infrastructure in Italy

Officially documented AI infrastructure and API coverage for Italy. A service operating nearby does not prove that a particular model runs here; endpoint, request-storage, inference, and routing claims remain separate.

Verified provider coverage

Inference, request storage, and endpoint coverage are separate claims. Model-level entries are grouped so a regional matrix stays scannable.

Local amounts are ECB reference conversions, not provider-billed currency or tax-inclusive prices (Aug 26, 2026).

Amazon Bedrock

28 model-specific in-region claims are documented.

28 models
View 28 documented model-specific entries
  • Titan Text Embeddings V2 · in region
  • Gemma 3 27B PT · in region
  • Gemma 3 12B IT · in region
  • Gemma 3 4B IT · in region
  • MiniMax M2 · in region · $0.35/MTok in / $1.41/MTok out · ≈ €0.30 in (ECB reference)
  • MiniMax M2.1 · in region · $0.36/MTok in / $1.44/MTok out · ≈ €0.31 in (ECB reference)
  • MiniMax M2.5 · in region · $0.36/MTok in / $1.44/MTok out · ≈ €0.31 in (ECB reference)
  • Magistral Small 2509 · in region
  • Voxtral Small 24B 2507 · in region
  • Voxtral Mini 3B 2507 · in region
  • Ministral 14B 3.0 · in region
  • Ministral 3 8B · in region
  • Ministral 3B · in region
  • Devstral 2 123B · in region
  • NVIDIA Nemotron Nano 12B v2 VL BF16 · in region
  • NVIDIA Nemotron Nano 9B v2 · in region
  • Nemotron Nano 3 30B · in region
  • NVIDIA Nemotron 3 Super 120B · in region
  • GPT OSS Safeguard 120B · in region
  • gpt-oss-120b · in region · $0.20/MTok in / $0.79/MTok out · ≈ €0.17 in (ECB reference)
  • GPT OSS Safeguard 20B · in region
  • gpt-oss-20b · in region
  • Qwen3 235B A22B 2507 · in region
  • Qwen3 VL 235B A22B · in region
  • Qwen3 Next 80B A3B · in region
  • Qwen3 32B · in region
  • Qwen3-Coder-30B-A3B-Instruct · in region
  • GLM 4.7 Flash · in region

Official evidence: AWS Bedrock model regional availability