Weights and biases pricing

On “weights and biases pricing”, the UPROUTER.ONLINE directory tracks 1 listing for Weights & Biases — 1 with a free tier, 0 live in the latest status check.

Pricing rows are normalized to per-1M-token USD where a provider publishes rates; free allowances are converted at the provider's own pay-as-you-go rates and flagged as estimates. Figures that could not be verified are shown as “not quantified” rather than dropped or guessed.

1 of the 1 tracked listing run a free tier. The tracked value is the allowance converted at the provider's own rates — a signup bonus, a monthly quota or a usage cap — so the column compares allowances, not lifetime spend.

Pay-per-use rows show the minimum input and output rates per 1M tokens in the directory. Where a rate could not be verified against the provider’s page, the field is left blank rather than estimated.

Each of the 1 entries carries a community risk rating (low / medium / high / unrated) with the evidence behind it, a data-confidence score, and a last-verified date per field. Status is probed continuously; the latest result is shown per entry.

Treat the rating as a starting point, not a clearance: the reasoning is public on each entry, and the provider’s own terms always win. Where data is stale or unverified, the directory marks it instead of hiding it.

Start from what you actually need: a free allowance, a stable paid endpoint, or both. 1 of the 1 matching entries carry a free tier, so for most “weights and biases pricing” use cases there is a $0 starting point in the table above.

Narrow by risk tier and the last-verified date, then open the entry: the source URL, the terms and the status probe are the three things that matter. Ordering on this page is a data sort — free-credit value, then name — and nothing about it is sponsored. For Weights & Biases specifically, the rows above are the complete match set, not a sample.

1. Open the entry and read the source URL — the pricing, free-tier and access figures on the row come from the provider’s own page at that link.

2. Check the last-verified date and data-confidence score. Fresh, high-confidence rows are direct from the provider; older or community-reported rows are labeled as such. For Weights & Biases, compare the row against the provider's current pricing page before you commit.

3. Read the risk rating and its reasoning. A low tier means established operator, clear terms and a stable history — a high tier means at least one of those is missing.

4. Confirm the live status probe result and the access method (api-key, subscription, local) match how you intend to use it. Then review the provider’s terms yourself — inclusion here is informational, never an endorsement.

Predicate — published, non-internal entries whose name, domain or description contains “wandb” (case-insensitive LIKE)

Sort — free-credit value descending, nulls last, provider name ascending (disclosed data sort — never sponsored)

Freshness — counts and rows are live from the dataset (60-second server cache); each row keeps its own per-field last-verified date, and unverified or stale values are flagged in place

Limitations — we track published pricing, free-tier terms and status — not account-level behavior. Free-credit figures are normalized estimates; risk is a community evidence rating, never a guarantee; inclusion is informational only.

Full methodology, probe cadence and the correction pathway: about & methodology · raw data: public dataset (CC BY 4.0) · pricing and status on every row are measurements with dates, not guarantees.