OpenAlex published its 2026 roadmap recap on January 16, 2026, following its Q1 2026 town hall — a summary of what shipped and what’s coming across the underlying infrastructure, data model, and API business model. This is distinct from CASRAI’s evergreen OpenAlex overview guide and from the recent CASRAI piece on OpenAlex’s API-key mandate and per-call dollar pricing, which covers the pricing mechanics as they actually launched in February 2026. This post looks at the roadmap as a whole: the Walden infrastructure rewrite, Awards becoming a first-class graph entity, and how the January framing of “credit-based” pricing relates to the dollar-denominated system that followed it.
The Walden rewrite: from a slow monolith to Databricks
OpenAlex has confirmed that “the Walden rewrite is done” and that it “now runs on a modern Databricks infrastructure that lets us ship faster and iterate on data quality in days instead of months.” For a service that CRIS teams, institutional repositories, and bibliometrics tools increasingly depend on for bulk metadata (see CASRAI’s OpenAlex API guide for endpoint mechanics), the practical upshot is less about any single new field and more about release cadence: data-quality fixes and corpus expansions that used to take months of pipeline work can now ship in days.
The scale change accompanying the rewrite is substantial. OpenAlex added 192 million new works sourced from DataCite and repository content, bringing its total indexed works to roughly 477 million — described by OpenAlex as the largest connected repository of scholarship yet published. For downstream consumers, that means a meaningfully larger long-tail of repository-hosted and preprint-adjacent content than the corpus contained a year earlier, alongside the usual caveat that applies to any rapid corpus expansion: coverage growth and metadata-quality maturity don’t always arrive on the same timeline, so institutions doing assessment or reporting work should expect some settling period before treating the newly added records as fully deduplicated and disambiguated.
Awards becomes a first-class entity
The second major 2026 change is structural rather than infrastructural: OpenAlex has given grant awards their own first-class entity type and API endpoint, rather than representing funding only as scattered fields attached to individual Works records. Concretely, OpenAlex reports extracting 27 million funder links directly from full-text PDFs and integrating 15 new funders. This follows funding-metadata work OpenAlex had already been building out (see OpenAlex’s own “Funding metadata in OpenAlex” post for the earlier groundwork), and separate reporting has connected this expansion to a Wellcome-funded initiative to ingest grant data from 30-plus funders worldwide, including NIH, NSF, DOE, UKRI, Wellcome, ERC, DFG, ANR, SNSF, NSERC, CIHR, KAKEN/JSPS, ARC, and FAPESP — though CASRAI has not independently verified the exact grant figure or funder count against OpenAlex’s own primary documentation, and readers relying on the precise list should confirm directly against OpenAlex’s current entity documentation before citing it.
Why this matters structurally: an Award-as-entity model means a grant can be queried, linked, and curated as its own node in the graph — connected to the funder, the awarded institution, and every Work that acknowledges it — instead of being reconstructed after the fact from free-text acknowledgement strings on individual papers. That’s a meaningfully different data model for anyone building funder-attribution reporting, grant-outcome tracking, or research-funding intelligence on top of OpenAlex, and it brings OpenAlex’s data model a step closer to the kind of structured funder/award linkage that CRIS platforms and identifier registries have long tried to establish through other means.
Credit-based pricing, and how it relates to the pricing CASRAI already covered
OpenAlex’s January 2026 roadmap post frames its 2026 API monetization plan in credit terms: a singleton lookup (e.g. a single work by ID) costs 1 credit, a list or filter query costs 10 credits, PDF full-text content costs 100 credits, and vector search — described at the time as “coming soon” — costs 1,000 credits. Read on its own, this is a relative-cost framework: it tells you that a bulk list call is roughly ten times more expensive than a single lookup, and that PDF retrieval is far more expensive again, without necessarily fixing a dollar amount to a credit.
By the time OpenAlex’s usage-based pricing actually launched, in a separate late-February 2026 announcement, the terms had shifted to explicit dollar figures rather than an abstract credit unit: roughly $0.0001 per list/filter call, $0.001 per search call, and $0.01 per full-text PDF/XML download, with a single ID/DOI lookup free of charge and each API key receiving roughly $1 of free usage per day. CASRAI covers that launched system — including the newly mandatory API keys, the free daily allowance, and the paid Member/Member+/Supporter annual tiers arranged directly with OurResearch — in a dedicated piece: OpenAlex’s API-key mandate and usage-based pricing. Treat the January credit framing as the roadmap-stage announcement of the pricing shift and the February piece as the mechanics that actually shipped; OpenAlex’s own roadmap post does not specify an exchange rate between “credits” and dollars, so this piece does not attempt to reconcile the two into a single conversion figure.
Other roadmap items from the same update
- Affiliation-matching curation (targeted for February 2026): institutions gain the ability to edit the algorithm OpenAlex uses to match author affiliations to institution records, with corrections propagating within roughly a day rather than waiting on the next full reprocessing cycle.
- Author name disambiguation rewrite (targeted for end of Q1 2026): OpenAlex describes a full rewrite of its author disambiguation system, calling author disambiguation “the hardest problem in bibliometrics” — a long-standing pain point for any CRIS or ORCID-adjacent workflow that relies on OpenAlex author IDs.
- A sync service: intended to let integrators pull daily batch updates instead of continuously polling the API for changes, reducing both integration complexity and API-call volume (relevant given the new per-call pricing above).
- Institutional membership tiers at $5,000 and $20,000 annually, bundling data curation support, consulting, and enhanced API access on top of the core free/paid API pricing.
What this means for CRIS and identifier-ecosystem teams
None of these changes alter OpenAlex’s core commitments — it remains a free, CC0-licensed, fully open scholarly metadata catalog, distinct in that respect from subscription databases like Scopus and Web of Science (see CASRAI’s Scopus vs Web of Science vs OpenAlex comparison). What changes is the operational shape of using it at scale. Larger, faster-refreshing corpus data means bulk-metadata consumers should expect more frequent, smaller updates rather than infrequent large ones. Awards-as-entity means anyone building funder-attribution or grant-outcome dashboards on OpenAlex now has a real structured object to query instead of parsing acknowledgement text. And usage-based pricing — whichever unit it’s denominated in — means integrators who previously treated the API as unmetered need to budget call volume the same way they would for any metered vendor API, particularly for list/filter-heavy and PDF-retrieval workloads. Teams evaluating OpenAlex for research assessment or discovery use cases should also see CASRAI’s guide to OpenAlex’s research-assessment and discovery use cases for how these changes intersect with citation-based analysis specifically.
Sourcing and what remains unconfirmed
The Walden-rewrite completion, the 477-million-work / 192-million-new-work figures, the Awards-entity description (27 million funder links, 15 new funders), the January credit-pricing figures, the affiliation-matching and author-disambiguation timelines, the sync service, and the membership-tier pricing are all drawn directly from OpenAlex’s own January 16, 2026 roadmap post on its official blog. The Wellcome-funding figure and the 30-plus-funder list for Awards ingestion come from secondary reporting CASRAI has not independently cross-checked against OpenAlex’s own documentation, and OpenAlex’s roadmap post does not state a firm cutover date for when credit-denominated pricing became the dollar-denominated pricing covered in CASRAI’s dedicated pricing piece — readers with load-bearing dependencies on any of these specifics should verify directly against OpenAlex’s current blog and API documentation before relying on them operationally.
