Cheaper Inference

Cheaper Inference is an OpenAI-compatible LLM gateway. One API key gives access to models from several labs.

Setup

go get github.com/zendev-sh/goai@latest

import "github.com/zendev-sh/goai/provider/cheaperinference"

Set the CHEAPER_INFERENCE_API_KEY environment variable, or pass WithAPIKey() directly. Create a key at cheaperinference.com.

Models

Model IDs are bare names:

• gpt-5.4-mini
• gpt-5.4
• claude-sonnet-5
• deepseek-v4-flash
• glm-5.3

GET https://api.cheaperinference.com/public/models lists the full catalog and needs no API key. See cheaperinference.com/docs for details.

Tested Models

Unit tested (mock HTTP server): gpt-5.4-mini

Usage

model := cheaperinference.Chat("gpt-5.4-mini")

result, err := goai.GenerateText(ctx, model, goai.WithPrompt("Hello"))
if err != nil {
    log.Fatal(err)
}
fmt.Println(result.Text)

Options

Option	Type	Description
WithAPIKey(key)	string	Set a static API key
WithTokenSource(ts)	provider.TokenSource	Set a dynamic token source
WithBaseURL(url)	string	Override the default https://api.cheaperinference.com/v1 endpoint
WithHeaders(h)	map[string]string	Set additional HTTP headers
WithHTTPClient(c)	*http.Client	Set a custom *http.Client

Notes

• Declares text-only chat capability with tool calling. Image parts are sent in the standard image_url format for models that accept them.
• Streaming requests set stream_options.include_usage so the final chunk carries token usage.
• The package provides Chat only. Cheaper Inference also serves /embeddings (for example text-embedding-3-small), which provider/compat can reach with the same base URL.
• Environment variable CHEAPER_INFERENCE_BASE_URL can override the default endpoint.