🛡️ AgentFlare Docs

Supported Models

Pricing table for all models AgentFlare can track costs for.

AgentFlare has a hardcoded pricing table for the models below. Cost is calculated as:

cost = (input_tokens / 1_000_000) × input_price_per_1M
     + (output_tokens / 1_000_000) × output_price_per_1M

All prices are in USD.


OpenAI

ModelInput / 1M tokensOutput / 1M tokens
gpt-4o$2.50$10.00
gpt-4o-mini$0.15$0.60
gpt-4-turbo$10.00$30.00

Anthropic Claude

ModelInput / 1M tokensOutput / 1M tokens
claude-opus-4-6$15.00$75.00
claude-sonnet-4-6$3.00$15.00
claude-3-5-sonnet$3.00$15.00
claude-3-5-haiku$0.80$4.00
claude-haiku-4-5$0.80$4.00

Google Gemini

ModelInput / 1M tokensOutput / 1M tokens
gemini-1.5-pro$1.25$5.00
gemini-1.5-flash$0.075$0.30
gemini-2.0-flash$0.10$0.40

Unknown models

If you send an event with a model name that's not in the table, AgentFlare falls back to GPT-4o pricing ($2.50 / $10.00). The event is still recorded — you'll just see a cost estimate rather than an exact figure.

If you're using a model that's not listed, open an issue or PR to add it to the pricing table in agentguard/backend/services/cost.py.

On this page