Ask
Launch

W&B adds ARIA, a new token-priced product, to its pricing page

Weights & Biases pricing

Weights & Biases added ARIA — a token-based-pricing product — as its own metered line on the pricing compare table, free for a limited time on Free and Pro and included on Enterprise. Core tiers and meters unchanged.

Before

Pricing metered on seats, storage ($0.03/GB), Weave ingestion ($0.10/MB), and per-token Serverless Inference.

After

Same core tiers (Free / Pro $60 / Enterprise) plus a new ARIA token-priced product line, free for a limited time on Free and Pro; refreshed Serverless Inference catalog (30+ models, to $4.40/1M output).

Weights & Biases added ARIA, a token-based-pricing product, as a distinct metered line on its pricing compare table (labeled “Token based pricing”). It is offered free “for a limited time” on the Free and Pro tiers and is included on Enterprise. The addition sits alongside a refreshed W&B Serverless Inference model catalog — now 30+ open-weight models spanning roughly $0.01/1M input (DeepSeek V4-Flash) up to $4.40/1M output (Z.AI GLM 5.2).

The core structure was unchanged this capture: Free ($0), Pro (starts at $60/mo), and custom Enterprise, still metered on storage ($0.03/GB overage), Weave data ingestion ($0.10/MB overage), and per-token Inference, with CoreWeave Sandboxes credits ($10/$25 per month).

From Weights & Biases's pricing timeline
ARIA token-priced product added to the pricing page

W&B added ARIA — a new token-based-pricing product — as a distinct metered line on the pricing compare table, free 'for a limited time' on Free and Pro and included on Enterprise, alongside a refreshed Serverless Inference model catalog. Core tiers (Free / Pro $60 / Enterprise) and the storage ($0.03/GB) + Weave ingestion ($0.10/MB) meters were unchanged.

About Weights & Biases
wandb.ai ↗

Weights & Biases (W&B) is the MLOps experiment-tracking standard plus W&B Weave (LLM observability/evals), Models registry, and Serverless Inference — used by OpenAI, Meta, and Toyota across 1M+ users.

Free tier
Yes
Commits
Available
Transparency
public

Weights & Biases pricing history

  1. Aug 2026
    DeepSeek V4-Pro cut ~34%/27%; DeepSeek V4-Flash-0731 added
  2. Jul 2026
    Inference rate card repriced again; MiniMax M3 added
  3. Jul 2026
    DeepSeek V4-Flash repriced 14x on Serverless Inference
  4. Jul 2026
    ARIA token-priced product added to the pricing page
  5. Jan 2026
    Restructured tiers: storage + ingestion + per-token Inference
Full Weights & Biases timeline

More Weights & Biases activity

All pricing activity