tinker-training-cost

name	tinker-training-cost
description	Calculate training costs for Tinker fine-tuning jobs. Use when estimating costs for Tinker LLM training, counting tokens in datasets, or comparing Tinker model training prices. Tokenizes datasets using the correct model tokenizer and provides accurate cost estimates.

Tinker Training Cost Calculator

Calculate training costs for Tinker fine-tuning jobs by tokenizing your dataset with the correct model tokenizer and applying current pricing.

Quick Start

Use the bundled script to calculate training costs:

# List available models and pricing
python scripts/calculate_cost.py --list-models

# Calculate cost for a JSONL dataset (model matches on unambiguous fragment)
python scripts/calculate_cost.py training_data.jsonl --model Qwen3-8B --epochs 3

# Output as JSON
python scripts/calculate_cost.py training_data.jsonl --model Inkling --json

The script:

1. Loads the correct tokenizer for the selected model (via tinker-cookbook if installed, else transformers)
2. Counts tokens in your JSONL file (supports chat, text, and instruction formats)
3. Calculates the estimated training cost

Cost Formula

Training Cost = (total_tokens × epochs × train_price_per_million) / 1_000_000

---

Tinker Pricing

Prices effective July 17, 2026 (prefill/sample rose ~50%, train ~10% on that date) Source: https://tinker-docs.thinkingmachines.ai/tinker/models/

All prices in USD per million tokens. Prefill = input context (inference), Sample = output tokens (inference), Train = training tokens. Cached prefill tokens get an 80% discount. = extended-context variant.