TensorAPI.comTensor API
Define input meaning, numerical constraints, error states and output revisions so that clients can integrate with tensor workflows without guessing how each value is interpreted.
Explore this topicThe complete field guide
Twelve ways into
the world of tensors.
Choose a starting point, follow the data, and connect the concepts. Each guide focuses on a distinct question, with practical next steps and deeper reading from the Lab.
AI tensor
Match tensor layouts to text, image and tabular tasks, then document the preprocessing and output semantics that make those shapes meaningful throughout an AI workflow.
Explore this topicLLM Tensor API
Decide what an LLM integration should expose, understand common internal tensor roles, and return completion states and metadata that support real application workflows.
Explore this topicAI LLM Tensor API
Build a measured AI workflow with representative evaluation cases, explicit acceptance checks and task-level usage records, so each model or prompt change can be assessed.
Explore this topicAnthropic Tensor API
Connect Anthropic Messages to a wider AI pipeline with explicit responsibilities for numerical processing, evidence selection, response parsing and application-level validation.
Explore this topicAzure Tensor API
Match Azure endpoint choices to interactive or asynchronous inference, then define numerical schemas, package preprocessing and verify complete deployment behavior before rollout.
Explore this topicCursor Tensor API
Build API integrations through focused repository context, concise project rules, testable changes and careful review of adapter contracts and failure behavior.
Explore this topicTensor API prompts
Use a focused prompt recipe that joins the task, allowed evidence, output shape and validation checks, with explicit behavior when the source cannot support an answer.
Explore this topicTokenized Tensor API
Preserve source identities, observation times and transformation records when AI processes tokenized asset information, so every derived feature and classification remains traceable.
Explore this topicTokens Tensor API
Make token counts interpretable by recording tokenizer identity, complete request context, output allowances and observed usage separately from numerical tensor representations.
Explore this topicAI tokens Tensor API
Separate language model token accounting from blockchain asset records, then define clear field names, provider usage semantics and evidence requirements for each system.
Explore this topicClassification Tensor API
Define labels and review states, interpret model scores carefully, and choose thresholds from evaluated tradeoffs between correct automation, missed cases and manual review.
Explore this topic