Huggingface

Added Nvidia RTX PRO 6000 GPU hardware options (1x/2x/4x/8x) to Spaces hardware pricing at $2.75, $5.50, $11.00 and $22.00 per hour

PRO plan's included inference credits benefit changed from a relative multiplier ('20x included inference credits') to an explicit absolute monthly dollar allowance ('$2 of included inference credits monthly')

Added NVIDIA RTX PRO 6000 GPU instance option to Inference Endpoints (1/2/4/8 GPU configs, $2.75-$22.00/hour) Team plan SSO now explicitly supports OIDC in addition to SAML; Enterprise gained automated user management via SCIM provisioning

Quarterly roll-up · includes 4 weeklies

PRO section heading changed from 'Join 35k+ PRO members' to 'Do more with Hugging Face PRO', shifting from social proof to value proposition framing.

ZeroGPU Spaces hardware upgraded from Nvidia H200 (70 GB VRAM) to Nvidia RTX Pro 6000 Blackwell (up to 96 GB VRAM); price remains FREE for all users.

NVIDIA RTX PRO 6000 GPU added to Inference Endpoints on AWS with 4 configurations: 1x96GB ($2.75/hr), 2x192GB ($5.50/hr), 4x384GB ($11.00/hr), 8x768GB ($22.00/hr).

Enterprise plan adds SCIM provisioning for automated user management and renames support to 'Dedicated'; Team plan SSO expanded to include OIDC alongside SAML.

New 'Data Storage' section added with per-TB pricing: Base $12/TB/mo (public) and $18/TB/mo (private). Volume discounts: 50TB+ 20% off, 200TB+ 25% off, 500TB+ 33% off (down to $8/TB/mo public).

Quarterly roll-up · includes 4 weeklies

Data Storage now supports Buckets alongside models, datasets, and Spaces

Added 2x public storage capacity to PRO Account plan

Added Advanced security and access controls feature to Enterprise plan

New Data Storage offering added with per-TB pricing: Base $12/TB/mo (public $18/TB/mo), volume discounts at 50TB+ (20% off), 200TB+ (25% off), 500TB+ (33% off, $8/TB/mo public).

Nvidia H100 hardware options removed from Spaces Hardware section - 1x H100 (23 vCPU, 240GB RAM, 80GB VRAM at $4.50/hr) and 8x H100 (184 vCPU, 1920GB RAM, 640GB VRAM at $36.00/hr) discontinued as Spaces options Spaces Persistent Storage options removed - Small (20GB at $5/mo), Medium (150GB at $25/mo), and Large (1TB at $100/mo) tiers discontinued

AWS NVIDIA H100 added to Inference Endpoints: 1x 80GB at $4.50/hour (previously only GCP H100 at $10.00/hour) NVIDIA B200 GPU added to Inference Endpoints with multiple configurations: 1x 256GB at $9.25/hour, 2x 512GB at $18.50/hour, 4x 1024GB at $37.00/hour, 8x 2048GB at $74.00/hour

AWS NVIDIA A100 Inference Endpoints (1x 80GB) decreased from $4.00/hour to $2.50/hour (37.5% decrease) 10x private storage capacity added to PRO account

TPU v5e 2x2 hardware option ($4.75/hr) removed from Spaces Hardware TPU v5e 2x4 hardware option ($9.50/hr) removed from Spaces Hardware

Nvidia A100 - large Spaces Hardware VRAM increased from 40 GB to 80 GB at same $4.00/hour price Pro Account Inference API feature enhanced from 'higher rate limits' to 'x20 higher rate limits on Serverless API'

Nvidia H100 removed from Spaces Hardware on-demand section (was $10.00/hr Custom tier with 24 vCPU, 250GB RAM, 80GB VRAM). H100 remains available in Inference Endpoints. Centralized token control and approval added to Enterprise Hub

4x Nvidia L4 hardware option added to Spaces at $3.80/hour (48 vCPU, 186GB, 96GB VRAM) NVIDIA L4 Inference Endpoint option added at $0.80/hour (1 GPU, 24GB) on AWS

Priority support added to Enterprise Hub Granular access control with Resource groups added to Enterprise Hub