tokencost 0.2.10

Clientside token counting + price estimation for LLM apps and AI agents ported to Dart from Python package at https://github.com/AgentOps-AI/tokencost.

0.2.10

• Update prices and context window local file to Jun 30, 2024.
• Add TokenCost.calculateCostByDuration() to calculate the cost using the duration in seconds.

0.2.9

• Update prices and context window local file to May 21, 2024.
• Update all dependencies to their latest versions.

0.2.8

• Update prices and context window local file to May 09, 2024.
• Update all dependencies to their latest versions.

0.2.7

• Update prices and context window local file to May 04, 2024.

0.2.6

• Update prices and context window local file to Apr 02, 2024.

0.2.5

• Update prices and context window local file to Mar 19, 2024.

0.2.4

• Update prices and context window local file to Mar 18, 2024.
• Add all other possible fields in ModelPricesAndContextWindow
• Add TokenCost.calculateCostByPixels() to calculate the cost when pixel count is available

0.2.3

• Update prices and context window local file to Mar 14, 2024.

0.2.2

• Add TokenCost.calculateCostByTokens() to calculate the cost when token count is available

0.2.1

• Update prices and context window local file.

0.2.0

• using package:money2 instead of TPU units. This is a breaking change, as the API has changed.

0.1.1

• Converted model_prices.yaml file to a String constant, so it works in other packages

0.1.0

• Consider 0.0.6 invalid as according to https://dart.dev/tools/pub/versioning#semantic-versions we should have changed the minor version number as there is a breaking changed in the API.
• Breaking change in the API: 1 Token Price Unit (TPU) is now defined as 1/100,000,000 of $1 (USD). 1,000,000 TPUs would equate to $0.01.

0.0.6

• Update to latest changes in https://github.com/AgentOps-AI/tokencost/, except support for LlamaIndex callbacks
• Use LiteLLMs cost dictionary from https://github.com/BerriAI/litellm/blob/main/model_prices_and_context_window.json
• Add new pricing models as of 2024-01-26:
  • Including two new OpenAI embedding models: text-embedding-3-small, and text-embedding-3-large
  • New aliases gpt-3.5-turbo, gpt-4, and gpt-4-turbo-preview
  • An update of GPT-4 and GPT-3.5 prices
  • Included a lot of other LLMs, proprietary and open source

0.0.5

• Improve README.md and pubspec.yaml

0.0.4

• Add link to https://github.com/AgentOps-AI/tokencost in README.md

0.0.3

• Add platforms in pubspec.yaml

0.0.2

• Initial release of tokencost, full port as a Dart package of AgentOps-AI/tokencost under MIT license.