AI model API pricing
relay.fast gives one API key for GPT, Claude, Gemini, Grok, DeepSeek, GLM, Kimi, Qwen and other models through OpenAI, Anthropic and Gemini-compatible endpoints. The tables list the official price of each model next to the lowest current relay.fast price.
Prices checked from the live catalog. "From" is the lowest price among the relay.fast sources that currently serve the model. The price you pay depends on the source your request uses and can change on any day. The pricing page always shows current values.
Text and coding models
Anthropic
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| Claude Fable 5 | claude-fable-5 | $10 / $50 | $0.278 / $1.39 (−97%) | $0.0278 | 1M |
| Claude Fable 5.1 | claude-fable-5-1 | $10 / $50 | $2.23 / $11.16 (−77%) | $0.0558 | 1M |
| Claude Haiku 4.5 | claude-haiku-4-5 | $1 / $5 | $0.0278 / $0.139 (−97%) | $0.00278 | 200K |
| Claude Haiku 5.5 requests above 100K input tokens cost more | claude-haiku-5-5 | $0.1 / $0.5 | $0.0223 / $0.112 (−77%) | $0.00223 | 1M |
| Claude Opus 4.6 | claude-opus-4-6 | $5 / $25 | $0.139 / $0.694 (−97%) | $0.0139 | 1M |
| Claude Opus 4.7 | claude-opus-4-7 | $5 / $25 | $0.139 / $0.694 (−97%) | $0.0139 | 1M |
| Claude Opus 4.8 | claude-opus-4-8 | $5 / $25 | $0.139 / $0.694 (−97%) | $0.0139 | 1M |
| Claude Opus 5 | claude-opus-5 | $5 / $25 | $0.139 / $0.694 (−97%) | $0.0139 | 1M |
| Claude Opus 5.5 | claude-opus-5-5 | $4 / $20 | $0.111 / $0.556 (−97%) | $0.00556 | 1M |
| Claude Sonnet 4.6 | claude-sonnet-4-6 | $3 / $15 | $0.0833 / $0.417 (−97%) | $0.00833 | 1M |
| Claude Sonnet 5 | claude-sonnet-5 | $3 / $15 | $0.0833 / $0.417 (−97%) | $0.00833 | 1M |
| Claude Sonnet 5.5 | claude-sonnet-5-5 | $2 / $10 | $0.238 / $1.19 (−88%) | $0.0238 | 1M |
OpenAI
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| GPT-5.5 requests above 272K input tokens cost more | gpt-5.5 | $5 / $30 | $0.075 / $0.45 (−98%) | $0.0075 | 272K |
| GPT-5.6 Luna requests above 272K input tokens cost more | gpt-5.6-luna | $0.2 / $1.2 | $0.016 / $0.096 (−92%) | $0.0016 | 1M |
| GPT-5.6 Sol requests above 272K input tokens cost more | gpt-5.6-sol | $5 / $30 | $0.075 / $0.45 (−98%) | $0.0075 | 1M |
| GPT-5.6 Terra requests above 272K input tokens cost more | gpt-5.6-terra | $2 / $12 | $0.03 / $0.18 (−98%) | $0.003 | 1M |
| GPT-6 Astra requests above 272K input tokens cost more | gpt-6-astra | $10 / $50 | $0.1 / $0.5 (−99%) | $0.01 | 1.05M |
| GPT-6 Luna requests above 272K input tokens cost more | gpt-6-luna | $0.1 / $0.5 | $0.008 / $0.04 (−92%) | $0.0008 | 1.05M |
| GPT-6 Sol requests above 272K input tokens cost more | gpt-6-sol | $2 / $10 | $0.03 / $0.15 (−98%) | $0.003 | 1.05M |
| GPT-6.1 Sol requests above 272K input tokens cost more | gpt-6.1-sol | $2 / $10 | $0.03 / $0.15 (−98%) | $0.0015 | 1.05M |
| GPT-6.1 Terra requests above 272K input tokens cost more | gpt-6.1-terra | $1 / $5 | $0.034 / $0.17 (−96%) | $0.0017 | 1.05M |
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| Gemini 3.8 Flash | gemini-3.8-flash | $1.5 / $7.5 | $0.06 / $0.3 (−96%) | $0.006 | 1.05M |
| Gemini 4 Argon | gemini-4-argon | $4 / $20 | $0.32 / $1.6 (−92%) | $0.016 | 1.05M |
xAI
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| Grok 4.5 | grok-4.5 | $2 / $6 | $0.0333 / $0.1 (−98%) | $0.00833 | 500K |
| Grok 4.6 requests above 200K input tokens cost more | grok-4.6 | $2 / $6 | $0.0333 / $0.1 (−98%) | $0.00833 | 500K |
| Grok 4.7 requests above 200K input tokens cost more | grok-4.7 | $2 / $6 | $0.0133 / $0.04 (−99%) | $0.00333 | 500K |
DeepSeek
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| DeepSeek V4 Flash price depends on the time of day; the peak-hours price is shown | deepseek-v4-flash | $0.3 / $1.2 | $0.015 / $0.06 (−95%) | $0.0003 | 1M |
| DeepSeek V4 Flash 0731 price depends on the time of day; the peak-hours price is shown | deepseek-v4-flash-0731 | $0.3 / $1.2 | $0.015 / $0.06 (−95%) | $0.0003 | 1M |
| DeepSeek V4 Flash Vision Exp price depends on the time of day; the peak-hours price is shown | deepseek-v4-flash-vision-exp | $0.3 / $1.2 | $0.03 / $0.12 (−90%) | $0.0006 | 1M |
| DeepSeek V4 Pro price depends on the time of day; the peak-hours price is shown | deepseek-v4-pro | $1.32 / $3.96 | $0.066 / $0.198 (−95%) | $0.0022 | 1M |
| DeepSeek V4 Pro 0813 price depends on the time of day; the peak-hours price is shown | deepseek-v4-pro-0813 | $1.32 / $3.96 | $0.066 / $0.198 (−95%) | $0.0022 | 1M |
| DeepSeek V4.1 Flash price depends on the time of day; the peak-hours price is shown | deepseek-v4.1-flash | $0.3 / $1.2 | $0.03 / $0.12 (−90%) | $0.0006 | 1M |
| DeepSeek V4.1 Flash 1017 price depends on the time of day; the peak-hours price is shown | deepseek-v4.1-flash-1017 | $0.3 / $1.2 | $0.03 / $0.12 (−90%) | $0.0006 | 1M |
Zhipu AI
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| GLM-5.2 | glm-5.2 | $1.4 / $4.4 | $0.035 / $0.11 (−97%) | $0.0065 | 1M |
| GLM-5.3 | glm-5.3 | $1.4 / $4.4 | $0.035 / $0.11 (−97%) | $0.0065 | 1M |
| GLM-5.3 Flash | glm-5.3-flash | $0.15 / $0.5 | $0.0187 / $0.0625 (−87%) | $0.00375 | 1M |
Moonshot
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| Kimi K2.7 Code | kimi-k2.7-code | $0.95 / $4 | $0.095 / $0.4 (−90%) | $0.019 | 256K |
| Kimi K3 | kimi-k3 | $3 / $15 | $0.15 / $0.75 (−95%) | $0.015 | 1.05M |
| Kimi K3.1 Pro | kimi-k3.1-pro | $2 / $10 | $0.2 / $1 (−90%) | $0.02 | 1.05M |
Alibaba
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| Qwen3.8 27b | qwen3.8-27b | $0.5 / $3 | $0.0167 / $0.1 (−96%) | $0.00333 | 1M |
| Qwen3.8 Flash | qwen3.8-flash | $0.15 / $0.47 | $0.0075 / $0.0235 (−95%) | $0.0008 | 1M |
| Qwen3.8 Max | qwen3.8-max | $2 / $6 | $0.0667 / $0.2 (−96%) | $0.00833 | 1M |
MiniMax
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| MiniMax M3 | minimax-m3 | $0.3 / $1.2 | $0.05 / $0.2 (−83%) | $0.01 | 1M |
| MiniMax M3.1 Flash | minimax-m3.1-flash | $0.6 / $2.4 | $0.1 / $0.4 (−83%) | $0.02 | 1M |
| MiniMax M3.1 Flash Preview | minimax-m3.1-flash-preview | $0.6 / $2.4 | $0.1 / $0.4 (−83%) | $0.02 | 1M |
Xiaomi
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| MiMo V2.6 Pro | mimo-v2.6-pro | $0.435 / $0.87 | $0.0725 / $0.145 (−83%) | $0.0006 | 1M |
Tencent
| Model | Model ID | Official input / output per 1M tokens | relay.fast from, input / output | Cache read from | Context |
|---|---|---|---|---|---|
| Hy3 | hy3 | $0.132 / $0.528 | $0.0198 / $0.0792 (−85%) | $0.00495 | 256K |
Image generation models
| Model | Model ID | List price | Price from |
|---|---|---|---|
| GPT Image 2 | gpt-image-2 | $0.5 per image | $0.01 per image (−98%) |
| GPT Image 2.5 | gpt-image-2.5 | $0.5 per image | $0.01 per image (−98%) |
| GPT Image 2.5 Flare | gpt-image-2.5-flare | $0.5 per image | $0.01 per image (−98%) |
| GPT Image 2.5 Flare 2K | gpt-image-2.5-flare-2k | $3.75 per image | $0.075 per image (−98%) |
| GPT Image 2.5 Flare 4K | gpt-image-2.5-flare-4k | $3.75 per image | $0.075 per image (−98%) |
| GPT Image 2.5 Sunburst | gpt-image-2.5-sunburst | $0.5 per image | $0.01 per image (−98%) |
| GPT Image 2.5 Sunburst 2K | gpt-image-2.5-sunburst-2k | $3.75 per image | $0.075 per image (−98%) |
| GPT Image 2.5 Sunburst 4K | gpt-image-2.5-sunburst-4k | $3.75 per image | $0.075 per image (−98%) |
| GPT Image 2.6 | gpt-image-2.6 | $0.5 per image | $0.01 per image (−98%) |
| GPT Image 2.6 Flare | gpt-image-2.6-flare | $0.5 per image | $0.01 per image (−98%) |
| GPT Image 2.6 Sunburst | gpt-image-2.6-sunburst | $0.5 per image | $0.01 per image (−98%) |
| Gemini Nano Banana 2.1 | gemini-nano-banana-2.1 | $0.6 per image | $0.0375 per image (−93%) |
| Grok Imagine Image 2.0 | grok-imagine-image-2.0 | $1.2 per image | $0.04 per image (−96%) |
Video generation models
| Model | Model ID | List price | Price from |
|---|---|---|---|
| Gemini Omni Edit | gemini-omni-edit | $0.29 per video | $0.29 per video |
| Gemini Omni Fast | gemini-omni-fast | $0.235 per video | $0.235 per video |
| Veo 3.1 Lite | veo-3.1-lite | $0.12 per second of video | $0.12 per second of video |
| Grok Imagine Video 1.5 | grok-imagine-video-1.5 | $0.036 per second of video | $0.036 per second of video |
| Wan 3.0 | wan-3.0 | $0.125 per second of video | $0.125 per second of video |
| MiniMax H3 768p | minimax-h3-768p | $0.015 per second of video | $0.015 per second of video |
| Seedance 2.0 Fast | seedance-2.0-fast | $0.061 per second of video | $0.061 per second of video |
| Seedance 2.0 Mini | seedance-2.0-mini | $0.037 per second of video | $0.037 per second of video |
| Seedance 2.5 | seedance-2.5 | $0.27 per second of video | $0.27 per second of video |
| Kling 3.0 | kling-3.0 | $0.175 per second of video | $0.175 per second of video |
| Vidu Q3 Turbo | viduq3-turbo | $0.165 per second of video | $0.165 per second of video |
Text to speech models
| Model | Model ID | List price | Price from |
|---|---|---|---|
| Grok Voice | grok-voice | $0.032 per request | $0.032 per request |
| Chirp 3 HD | chirp-3-hd | $0.0096 per request | $0.0096 per request |
How the prices work
- The official price is the model developer's list price recorded in the catalog, per 1M input and output tokens.
- The relay.fast price depends on the source a request uses. "From" is the lowest price among the sources that serve the model now.
- Cache reads are billed at the cache price when the model reports cached input tokens.
- Some models charge more for long inputs or at peak hours; such rows carry a note.
- Requests are charged to the account balance as they run. There is no subscription.
Machine-readable pricing
The same catalog is available as JSON at https://relay.fast/pricing.json. A summary for AI assistants is at llms.txt. Setup guides for coding tools are in the API docs.
Frequently asked questions
How are relay.fast API prices calculated?
Each model has an official price per 1M input and output tokens, or per image, second, or request for media models. relay.fast charges a fraction of it that depends on the source your request uses. The tables show the lowest current price, checked 2026-11-05 15:22 UTC.
Is there a subscription or minimum spend?
relay.fast is pay-as-you-go: requests are charged to the account balance by the tokens or units they use. Read the current terms before buying.
Where can a program read current prices?
GET https://relay.fast/pricing.json returns the live catalog with model prices, groups, endpoints, and sources.
Can prices change?
Yes. Prices, models, and sources can change on any day. Check the pricing page before production use instead of relying on an older copy.
How much does Claude Opus 5.5 cost on relay.fast?
From $0.111 per 1M input tokens and $0.556 per 1M output tokens, against the official $4 and $20 (checked 2026-11-05).