Thinking Machines Lab

Thinking Machines Lab is an AI research company (founded by Mira Murati) whose first product is Tinker — a cloud training API for researchers and developers. Tinker gives you full control over your data, algorithms, and models while it manages the GPU infrastructure: you author training loops locally with four primitives (forward_backward, optim_step, sample, save_state) and Tinker runs distributed LoRA fine-tuning across dense and Mixture-of-Experts models from 1B to 1T+ parameters. It supports SFT, reinforcement learning (GRPO, PPO), DPO, and distillation, shipped as an official Python SDK (tinker) plus the tinker-cookbook of recipes, a web console/playground, and OAuth 2.0 / OIDC authentication.

Thinking Machines Lab is profiled on the APIs.io network. Tagged areas include Company, Artificial Intelligence, Machine-Learning, LLM, and Fine-Tuning.

Thinking Machines Lab’s developer surface includes documentation, API reference, getting-started guide, engineering blog, support, pricing, signup flow, and 23 more developer resources.

36.0/100 thin ▬ flat Agent 15/100 agent aware Full breakdown ↓
scored 2026-08-20 · rubric v0.12.0
AccessSelf serve
0 APIs
CompanyArtificial IntelligenceMachine-LearningLLMFine-TuningModel TrainingDeveloper ToolsReinforcement Learning

Kin Score

Kin Score Kin Score How this is scored →
scored 2026-08-20 · rubric v0.12.0
Composite quality — 36.0/100 · thin
Contract Quality 0.0 / 25
Developer Ergonomics 12.9 / 20
Access Clarity 8.9 / 20
Operational Transparency 6.8 / 13
Contract Governance 0.5 / 12
Discoverability 6.9 / 10
Agent readiness — 15/100 · agent aware
Machine-Readable Contract 0 / 18
Agentic Access Contract 0 / 10
Documented Reversibility 0 / 6
MCP Server 0 / 12
Machine-Readable Auth 10 / 10
Idempotency 0 / 9
Stable Error Semantics 8 / 8
Request/Response Examples 0 / 7
Rate-Limit Signaling 0 / 7
Typed Event Surface 0 / 6
Agent Skills 0 / 5
Well-Known Catalog 0 / 4
Consent & Bot Identity 3 / 3
A2A Agent Card 0 / 8
Dry-Run / Simulate Mode 0 / 4
Improve this rating by publishing the missing artifacts — every area above can be raised, and the full rubric is at apis.io/rating/. This rating is computed from github.com/api-evangelist/thinking-machines-lab: open an issue to ask a question, or submit a pull request to add artifacts. Want it done for you? Prioritized profiling — $2,500 →

Security Posture 3

Authentication, domain security, vulnerability disclosure, and trust-center signals.

Thinking Machines Lab Authentication

oauth2/openIdConnect · 2 schemes

SECURITY

Thinking Machines Lab Domain Security

TLSv1.3 · DMARC

SECURITY

Thinking Machines Lab Vulnerability Disclosure

security.txt · contact published

SECURITY

Scopes 1

OAuth scopes governing access to this provider's APIs.

Thinking Machines Lab Scopes

4 scopes · authorizationCode/deviceCode

4 scopes

SCOPES

Resources

Get Started 3

Portal, sign-up, and the first successful call

Documentation 2

Reference material describing how the API behaves

Agent Surfaces 2

MCP servers, agent skills, and machine-readable catalogs

Design & Contract 5

Pagination, idempotency, versioning, errors, and events

Build 3

SDKs, sample code, and the tooling you integrate with

Access & Security 6

Authentication, authorization, and security posture

Operate 4

Status, limits, changes, and where to get help

Commercial 3

Pricing, plans, and the legal terms of use

Company 2

The organization behind the API

Source (apis.yml)

apis.yml Raw ↑
aid: thinking-machines-lab
name: Thinking Machines Lab
description: 'Thinking Machines Lab is an AI research company (founded by Mira Murati) whose first product is Tinker — a cloud
  training API for researchers and developers. Tinker gives you full control over your data, algorithms, and models while
  it manages the GPU infrastructure: you author training loops locally with four primitives (forward_backward, optim_step,
  sample, save_state) and Tinker runs distributed LoRA fine-tuning across dense and Mixture-of-Experts models from 1B to 1T+
  parameters. It supports SFT, reinforcement learning (GRPO, PPO), DPO, and distillation, shipped as an official Python SDK
  (tinker) plus the tinker-cookbook of recipes, a web console/playground, and OAuth 2.0 / OIDC authentication.'
url: https://raw.githubusercontent.com/api-evangelist/thinking-machines-lab/refs/heads/main/apis.yml
accessModel:
  pricing: unknown
  onboarding: self-serve
  trial: false
  try_now: false
  public: false
  label: Self-serve signup
  confidence: medium
  source:
  - authentication
  generated: '2026-07-22'
  method: derived
image: https://thinkingmachines.ai/images/apple-touch-icon.png
x-type: company
x-source: vc-portfolio
x-backed-by:
- gv
x-tier: stub
x-tier-reason: portfolio-lead
specificationVersion: '0.23'
created: '2026-07-17'
modified: '2026-07-21'
tags:
- Company
- Artificial Intelligence
- Machine-Learning
- LLM
- Fine-Tuning
- Model Training
- Developer Tools
- Reinforcement Learning
tags_raw:
- Company
- Ai
- Machine Learning
- LLM
- Fine Tuning
- Model Training
- Developer Tools
- Reinforcement Learning
apis: []
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com
- FN: APIs.json
  email: info@apis.io
common:
- type: Website
  url: https://thinkingmachines.ai
- type: DeveloperPortal
  url: https://tinker-docs.thinkingmachines.ai/
- type: Documentation
  url: https://tinker-docs.thinkingmachines.ai/
- type: APIReference
  url: https://tinker-docs.thinkingmachines.ai/tinker/api-reference/restclient/
- type: GettingStarted
  url: https://tinker-docs.thinkingmachines.ai/cookbook/quickstart/
- type: Blog
  url: https://thinkingmachines.ai/blog/
- type: GitHubOrganization
  url: https://github.com/thinking-machines-lab
- type: Support
  url: https://github.com/thinking-machines-lab/tinker-feedback
- type: Pricing
  url: https://tinker-docs.thinkingmachines.ai/tinker/models/
- type: SignUp
  url: https://auth.thinkingmachines.ai/sign-up
- type: TermsOfService
  url: https://thinkingmachines.ai/legal/terms/
- type: PrivacyPolicy
  url: https://thinkingmachines.ai/legal/privacy/
- type: StatusPage
  url: https://status.thinkingmachines.ai
- type: ChangeLog
  name: Tinker changelog
  url: changelog/thinking-machines-lab-changelog.yml
- type: Deprecation
  url: https://tinker-docs.thinkingmachines.ai/tinker/models/
- type: Packages
  url: packages/thinking-machines-lab-packages.yml
- type: SDKs
  url: packages/thinking-machines-lab-packages.yml
- type: Authentication
  url: authentication/thinking-machines-lab-authentication.yml
- type: OAuthScopes
  url: scopes/thinking-machines-lab-scopes.yml
- type: WellKnown
  url: well-known/thinking-machines-lab-well-known.yml
- type: SecurityTxt
  url: well-known/thinking-machines-lab-security.txt
- type: Conformance
  url: conformance/thinking-machines-lab-conformance.yml
- type: ErrorCatalog
  url: errors/thinking-machines-lab-error-codes.yml
- type: Lifecycle
  url: lifecycle/thinking-machines-lab-lifecycle.yml
- type: Conventions
  url: conventions/thinking-machines-lab-conventions.yml
- type: DataModel
  url: data-model/thinking-machines-lab-data-model.yml
- type: LLMsTxt
  url: llms/thinking-machines-lab-llms.txt
- type: DomainSecurity
  url: security/thinking-machines-lab-domain-security.yml
- type: VulnerabilityDisclosure
  url: security/thinking-machines-lab-vulnerability-disclosure.yml
- type: Security
  url: https://thinkingmachines.ai/security/disclosure-policy/
x-enrichment:
  date: '2026-07-21'
  status: enriched
  artifacts_added: 16
  pass: local-v1