HeyTensor

What is HeyTensor?

HeyTensor is a set of free PyTorch tools by a solo developer, Michael Lip, that run entirely in your browser. The calculators work out output shapes for layers and operations such as F.pad, Conv3d, adaptive pooling, embeddings, LayerNorm, multi-head attention and ViT patch embedding, and also cover reshape and view, receptive fields, strides and parameter counts. A shape mismatch debugger takes a pasted PyTorch error and returns a plain English explanation with a fix, and guides walk through common errors like mat1 and mat2 shapes cannot be multiplied. Calculator inputs are never sent to a server. For people debugging their own models, HeyTensor also sells Tensor Preflight, a downloadable toolkit that traces a local forward pass, writes HTML and JSON reports and turns a repair into a regression check. It costs $29 as a one-time payment, needs Python 3.9+ and PyTorch 2.8+, and runs locally without telemetry. The calculators and guides stay free.

Use cases

• Calculate PyTorch layer output shapes
• Explain pasted tensor shape errors
• Count parameters of a network
• Check loss target shapes and dtypes
• Trace a local forward pass
• Keep shape checks beside your code

Who it's for

• ML engineers
• PyTorch developers
• Researchers and students

Why use HeyTensor

• The browser calculators answer layer-size questions instantly without a sign-up or server call.
• Pasting an error message gives a readable cause and a fix, which cuts time spent decoding PyTorch RuntimeErrors.
• The local toolkit keeps your model on your machine, with no toolkit telemetry or model uploads.
• A $29 one-time payment avoids a subscription for the paid toolkit.
• A sample report is available before buying so you can check the output format.

HeyTensor features

• Output shape calculators for Conv3d, adaptive pooling and F.pad
• Embedding, LayerNorm and MultiHead Attention shape calculators
• Shape mismatch error debugger with plain English fixes
• Reshape and view calculator
• CNN receptive field and ViT patch embedding calculators
• Neural network parameter counter
• Loss target shape checker for 8 loss functions
• Tensor Preflight forward-pass tracing with HTML and JSON reports
• Eight broken and fixed PyTorch workflows
• Reusable output contracts for shape, dtype and device

HeyTensor integrations

• PyTorch
• GitHub

How to get started with HeyTensor

1. Open a calculator on the site, enter your input shape and layer settings, and read the output shape.
2. For an error you cannot decode, paste it into the shape mismatch debugger to get an explanation and fix.
3. To debug a local model, buy Tensor Preflight, unzip it into your workspace and run it on a trusted model factory.
4. Open the generated HTML report, compare the broken and fixed variants, and save the expected shapes as a contract.

HeyTensor limitations

• Tensor Preflight needs Python 3.9+ and PyTorch 2.8+
• Tensor Preflight does not run your model on the website or repair it automatically
• Tracing observes one execution path, and module hooks miss some functional operations
• CUDA and broader accelerator compatibility are listed as untested

HeyTensor pricing

Yes, HeyTensor has a free plan. Paid plans start at $29 one-time, according to the official pricing page on 24 Oct 2026.

HeyTensor plans

• Free tools

  Free

  • Shape calculators
  • Error debugger
  • Guides and references
  • Runs in your browser

• Tensor Preflight

  $29

  one-time payment, ZIP download

  • Forward-pass tracing
  • HTML and JSON reports
  • Eight broken and fixed workflows
  • Runs locally

Plans as listed on the official pricing page on 24 Oct 2026. Prices can change and may vary by country or exclude tax. Check the official pricing page.

HeyTensor alternatives

The main HeyTensor alternatives listed on Debuly are Busabase, DVC, Inferable, LeanCTX, MemPalace and Middleware. The table compares pricing and platforms; order is A–Z.

1. 

   Busabase

   MIT-licensed database and workspace where AI agents store data, docs, skills and apps, available on desktop, self-hosted or in Busabase Cloud.

   Developer platformsAutomation and agentsProductivity
   Open sourceVisit Busabase (official site)
2. 

   DVC

   Open-source, Git-based data version control that tracks datasets, models and experiments next to your code in ML projects.

   Developer platformsCodingResearch and search
   Open sourceVisit DVC (official site)
3. In

   Inferable

   Durable workflow runtime for AI agents, with structured LLM output and workflows that run on your own compute.

   Developer platformsAutomation and agentsCoding
   FreemiumVisit Inferable (official site)
4. 

   LeanCTX

   Open-source context gateway that trims what AI coding tools read, caches repeat reads and logs what each agent was given.

   Developer platformsCodingAutomation and agents
   FreemiumVisit LeanCTX (official site)
5. 

   MemPalace

   Local-first AI memory that keeps your conversations as verbatim text and finds them again with semantic search through MCP.

   Developer platformsProductivityCoding
   Open sourceVisit MemPalace (official site)
6. 

   Middleware

   OpenTelemetry-native observability platform for infrastructure, APM, logs, RUM and LLM monitoring, with an AI agent for root cause analysis.

   Developer platformsAutomation and agentsCoding
   FreemiumVisit Middleware (official site)

HeyTensor FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from heytensor.com · fetched 23 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.