Models
Every model, at a fraction of its official price
Each model is priced per token at a fixed fraction of its maker's official price. The multiple says how many times less you pay: 5x less means one fifth of the official price. Open models carry their own, smaller multiple, shown on each row.
Check the official price on the maker's page: Anthropic, OpenAI, Google, Z.ai, Qwen (Alibaba) and DeepSeek.
Frontier models
- claude-opus-5-5 Anthropic · 200k context5x less
- Input /1M
- Official price
$4.00, Kontinent AI $0.80 - Output /1M
- Official price
$20.00, Kontinent AI $4.00 - Cached /1M
- Official price
$0.20, Kontinent AI $0.04
chat messages - claude-opus-5 Anthropic · 200k context5x less
- Input /1M
- Official price
$5.00, Kontinent AI $1.00 - Output /1M
- Official price
$25.00, Kontinent AI $5.00 - Cached /1M
- Official price
$0.50, Kontinent AI $0.10
chat messages - claude-sonnet-5-5 Anthropic · 200k context4x less
- Input /1M
- Official price
$2.00, Kontinent AI $0.50 - Output /1M
- Official price
$10.00, Kontinent AI $2.50 - Cached /1M
- Official price
$0.20, Kontinent AI $0.05
chat messages - claude-sonnet-5 Anthropic · 200k context5x less
- Input /1M
- Official price
$2.00, Kontinent AI $0.40 - Output /1M
- Official price
$10.00, Kontinent AI $2.00 - Cached /1M
- Official price
$0.20, Kontinent AI $0.04
chat messages - claude-haiku-4-5-20251001 Anthropic · 200k context5x less
- Input /1M
- Official price
$1.00, Kontinent AI $0.20 - Output /1M
- Official price
$5.00, Kontinent AI $1.00 - Cached /1M
- Official price
$0.10, Kontinent AI $0.02
chat messages - claude-fable-5-1 Anthropic · 200k context3x less
- Input /1M
- Official price
$10.00, Kontinent AI $3.30 - Output /1M
- Official price
$50.00, Kontinent AI $16.50 - Cached /1M
- Official price
$0.25, Kontinent AI $0.0825
chat messages - gpt-5.5 OpenAI · 1M context5x less
- Input /1M
- Official price
$5.00, Kontinent AI $1.00 - Output /1M
- Official price
$30.00, Kontinent AI $6.00 - Cached /1M
- Official price
$0.50, Kontinent AI $0.10
chat responses - gpt-6.1-sol OpenAI · 1M context5x less
- Input /1M
- Official price
$2.00, Kontinent AI $0.40 - Output /1M
- Official price
$10.00, Kontinent AI $2.00 - Cached /1M
- Official price
$0.10, Kontinent AI $0.02
chat responses - gpt-5 OpenAI · 400k context5x less
- Input /1M
- Official price
$1.25, Kontinent AI $0.25 - Output /1M
- Official price
$10.00, Kontinent AI $2.00 - Cached /1M
- Official price
$0.125, Kontinent AI $0.025
chat responses - gemini-3.1-pro-preview Google · 1M context5x less
- Input /1M
- Official price
$2.00, Kontinent AI $0.40 - Output /1M
- Official price
$12.00, Kontinent AI $2.40 - Cached /1M
- Official price
$0.20, Kontinent AI $0.04
chat - gemini-3.5-flash Google · 1M context5x less
- Input /1M
- Official price
$1.50, Kontinent AI $0.30 - Output /1M
- Official price
$9.00, Kontinent AI $1.80 - Cached /1M
- Official price
$0.15, Kontinent AI $0.03
chat
Open models
- glm-5.3 Z.ai · 1M context1.8x less
- Input /1M
- Official price
$1.40, Kontinent AI $0.77 - Output /1M
- Official price
$4.40, Kontinent AI $2.42 - Cached /1M
- Official price
$0.26, Kontinent AI $0.143
chat messages - qwen3.8-max Qwen (Alibaba) · 1M context1.8x less
- Input /1M
- Official price
$2.00, Kontinent AI $1.10 - Output /1M
- Official price
$6.00, Kontinent AI $3.30 - Cached /1M
- Official price
$0.25, Kontinent AI $0.1375
chat - deepseek-v4.1-flash DeepSeek · 1M context1.8x less
- Input /1M
- Official price
$0.30, Kontinent AI $0.165 - Output /1M
- Official price
$1.20, Kontinent AI $0.66 - Cached /1M
- Official price
$0.006, Kontinent AI $0.0033
chat messages
No model matches these filters.
USD per million tokens. The struck-through figure is the maker's official price. The figure beside it is ours. Formats: chat is /v1/chat/completions, messages is /v1/messages, responses is /v1/responses.
How prices work
- Our price is the official price times a fixed multiple. Input, output and cached input all use the same multiple.
- Cached input is priced at the same multiple of the maker's cached input price. Most of an agent's prompt is cached input.
- Where a maker charges more for a long prompt, our price for that prompt moves by the same step.
- Some models think before they answer. The thinking is billed as output tokens, at the same multiple.
- You pay only for requests that get an answer.
Check the numbers
GET https://api.kontinentai.com/v1/models returns the same prices, the official prices and the multiple for each model. It needs no API key.
The same table as data: the price index, with a CSV and a JSON download.
Every response also carries its own cost. kontinent.cost_usd is what
you paid. kontinent.list_price_usd is the same tokens at the official price.
kontinent.saved_usd is the difference.