Observed by Atlas — nobody has proved they control this domain. Is it yours?
TabPFN-3.5
OBSERVEDA tabular foundation model for prediction on structured data
- FRESHNESS
- 0.82
- DECAYS BETWEEN READS
- READINGS
- —
- VERIFIED / MEASURED
- VOUCHES
- 0
- SIGNED STATEMENTS
- THREADS
- 0
- FIELD NOTES
- CLICKS · 28D
- 0
- BOT-FILTERED
- FIRST SEEN
- SEP 18
- IN THE FLOW
TabPFN-3.5 is a foundation model for tabular data prediction, designed to handle messy, real-world business data such as insurance claims, transactions and sensor readings without extensive feature engineering or tuning. It is available via API, Private Cloud, on-prem deployment, SAP AI Core, AWS Sagemaker and as an open-source package. Variants include TabPFN-3.5-Plus for text-rich data, TabPFN-3.5-Thinking for highest accuracy, and TabPFN-3.5-Fast for latency-sensitive applications.
Nothing measured yet. A reading is a measurement with its evidence attached — uptime checked, pricing re-read against what the page claims. Raise a checkable claim in a field note below and Atlas will measure it.
Latest model, TabPFN-3.5-Plus, including Thinking mode · Free daily & monthly limits · Python SDK, REST API and MCP · Support from our team and community on Discord
Usage limits sized to your workload · Access to our technical support team · Faster inference with extended kv-cache · ...and more
Advanced security and procurement compliance · Dedicated support channel · Early access to new model versions · ...and more
Every crawl is diffed. Material changes are listed.
Nobody has vouched for TabPFN-3.5 yet.
A vouch says you use it. Your handle, how long, and an optional note.
I use this — vouchHave a question about TabPFN-3.5? Ask the people using it.
No field notes yet. A field note is about use: what broke, what it replaced, what the docs don't say. Anything in one that makes a checkable claim gets measured by Atlas, which replies with the reading.
Outbound clicks are counted for the builder.
Claim it