Pricing
Prices per 1M tokens: up to 10× cheaper than official
We have one price for the request and the response, and cache counts as regular tokens. You're charged after each request for actual usage, no subscription. Next to ours are the official prices from the model developers: they charge separately for input and output.
Example: a request to Claude Sonnet 5.5 of 20,000 tokens, including the response, costs $0.019.
| Model | API endpoints | Official, $ per 1M input / output | Slafy, $ per 1M input and output |
|---|---|---|---|
| Anthropic | |||
Claude Opus 5.5claude-opus-5-5 | chat/completions · responses · messages | $4.00 / $20.00 source | $1.43 |
Claude Sonnet 5.5claude-sonnet-5-5 | chat/completions · responses · messages | $2.00 / $10.00 source | $0.953 |
Claude Haiku 5.5claude-haiku-5-5 | chat/completions · responses · messages | $0.10 / $0.50 sourcefor requests up to 100K tokens | $0.278 |
Claude Fable 5.1claude-fable-5-1 | chat/completions · responses · messages | $10.00 / $50.00 source | $1.75 |
| OpenAI | |||
GPT-6.1 Solgpt-6.1-sol | chat/completions · responses · messages | $2.00 / $10.00 sourcefor context up to 272K tokens | $0.953 |
GPT-6.1 Terragpt-6.1-terra | chat/completions · responses | $1.00 / $5.00 sourcefor context up to 272K tokens | $0.318 |
GPT-6 Astragpt-6-astra | chat/completions · responses · messages | $10.00 / $50.00 sourcefor context up to 272K tokens | $1.27 |
GPT-6 Lunagpt-6-luna | chat/completions · responses · messages | $0.10 / $0.50 sourcefor context up to 272K tokens | $0.199 |
GPT Image 2.5 Sunburstgpt-image-2.5-sunburst | images | $8.00 / $30.00 sourceimage tokens | $0.794 |
Gemini 4 Argongemini-4-argon | chat/completions | $2.00 / $10.00 sourcetwice as expensive from December 1, 2026 | $1.59 |
Gemini 3.8 Flashgemini-3.8-flash | chat/completions | $0.75 / $3.75 sourcetwice as expensive from January 1, 2027 | $0.397 |
| xAI | |||
Grok 4.7grok-4.7 | chat/completions · responses · messages | $2.00 / $6.00 sourcefor requests up to 200K tokens | $0.636 |
| DeepSeek | |||
DeepSeek V4 Prodeepseek-v4-pro | chat/completions | $1.32 / $3.96 sourceat peak hours, half price the rest of the time | $0.318 |
DeepSeek V4.1 Flash 1017deepseek-v4.1-flash-1017 | chat/completions | $0.30 / $1.20 sourceat peak hours, half price the rest of the time | $0.119 |
| Moonshot AI | |||
Kimi K3.1 Prokimi-k3.1-pro | chat/completions | $2.00 / $10.00 source | $1.59 |
| Zhipu AI | |||
GLM 5.3glm-5.3 | chat/completions · responses · messages | $1.40 / $4.40 source | $0.318 |
GLM 5.3 Flashglm-5.3-flash | chat/completions · responses · messages | $0.15 / $0.50 source | $0.119 |
Official prices are taken from the developers' pricing pages, base rate without batch processing. Slafy prices in dollars are approximate, at the current exchange rate. Model errors and rejected requests aren't charged. If a request is interrupted before the model reports its exact usage, it's charged based on the volume of the request and the response received.