# Pricing transcribed from the provider's own page — see provenance. Numbers are quoted, not derived.
generated: '2026-10-07'
method: searched
source: https://tinker-docs.thinkingmachines.ai/tinker/models/models_and_pricing/
sources:
- https://tinker-docs.thinkingmachines.ai/tinker/models/models_and_pricing/
- https://tinker-docs.thinkingmachines.ai/tinker/models/
provenance:
  tool: all/0-working/economics/harvest_pricing.py
  run: 20261007T104522
  grounding: every price_text was found verbatim in the fetched page text
audience: developer-api
pricing_basis: per-call
delivery:
  delivery: managed
  method: page
  read: '2026-10-07'
published: true
note: Per-million-token price table for fine-tuning/inference across 29 models, plus beta serverless inference.
  80% discount on cached prefill tokens; several models carry limited-time 50% discounts. Only a representative
  subset of model rows transcribed as plans.
plan_count: 11
currency: USD
plans:
- name: Checkpoint storage
  price_text: $0.10 per GB per month
  price: 0.1
  currency: USD
  period: month
  unit: GB
- name: Inkling (64K) — prefill
  price_text: $1.87
  price: 1.87
  currency: USD
  period: per-use
  unit: per million tokens (prefill, limited-time 50% discount)
- name: Inkling (64K) — sample
  price_text: $4.68
  price: 4.68
  currency: USD
  period: per-use
  unit: per million tokens (sample)
- name: Inkling (64K) — train
  price_text: $5.61
  price: 5.61
  currency: USD
  period: per-use
  unit: per million tokens (train)
- name: Qwen3-8B — prefill
  price_text: $0.195
  price: 0.195
  currency: USD
  period: per-use
  unit: per million tokens (prefill)
- name: Qwen3-8B — sample
  price_text: $0.60
  price: 0.6
  currency: USD
  period: per-use
  unit: per million tokens (sample)
- name: Qwen3-8B — train
  price_text: $0.44
  price: 0.44
  currency: USD
  period: per-use
  unit: per million tokens (train)
- name: Serverless Inference (Beta) — Inkling input
  price_text: $1.00
  price: 1
  currency: USD
  period: per-use
  unit: per million input tokens
- name: Serverless Inference (Beta) — Inkling output
  price_text: $4.05
  price: 4.05
  currency: USD
  period: per-use
  unit: per million output tokens
- name: Serverless Inference (Beta) — Inkling-Small input
  price_text: $0.30
  price: 0.3
  currency: USD
  period: per-use
  unit: per million input tokens
- name: Serverless Inference (Beta) — Inkling-Small output
  price_text: $1.20
  price: 1.2
  currency: USD
  period: per-use
  unit: per million output tokens
x-canonical:
  generated: '2026-10-08'
  vocabulary: all/0-working/economics/plans-canonical.yml
  models:
  - hybrid
  units:
  - data
  - token
  plans:
  - name: Checkpoint storage
    access: other
    model: hybrid
    model_how: derived
    units:
    - data
    periods:
    - month
  - name: Inkling (64K) — prefill
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other
  - name: Inkling (64K) — sample
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other
  - name: Inkling (64K) — train
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other
  - name: Qwen3-8B — prefill
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other
  - name: Qwen3-8B — sample
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other
  - name: Qwen3-8B — train
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other
  - name: Serverless Inference (Beta) — Inkling input
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other
  - name: Serverless Inference (Beta) — Inkling output
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other
  - name: Serverless Inference (Beta) — Inkling-Small input
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other
  - name: Serverless Inference (Beta) — Inkling-Small output
    access: other
    model: hybrid
    model_how: derived
    units:
    - token
    periods:
    - other