ai() LLM Models

Use ai() to create provider clients and keep model traffic behind one Ax request shape.

auto client = axllm::OpenAICompatibleClient({
  {"api_key", api_key},
  {"model", "gpt-4.1-mini"},
});

What It Does

ai() selects a provider implementation from configuration and returns a client that Ax programs can call. The client handles chat, streaming, embeddings, media where supported, usage normalization, provider options, model keys, routing hooks, tracing, and runtime defaults.

flowchart LR
  A["Model key or alias"] --> B["Model catalog"]
  B --> C["Capability filter"]
  C --> D["Provider client"]
  D --> E["Request mapping"]
  E --> F["Provider API"]
  F --> G["Response normalization"]
  G --> H["Usage + trace"]

Core Call Shape

Create the client once near the application boundary, then pass it into forward(), streamingForward(), agents, flows, or optimizers.

client = ai(provider options)
result = program.forward(client, inputs)

Common Patterns

• Use a provider name and environment-backed API key.
• Set a default model in provider config when the app has one obvious model.
• Define model aliases when callers should choose fast, smart, or cheap instead of provider model IDs.
• Use OpenAI-compatible apiURL for compatible providers.
• Use model catalog helpers before runtime when the UI needs provider/model selectors.
• Use routers or balancers when provider fallback is part of the product.

Vertex routing and OpenAI prompt caching

Gemini and Anthropic Vertex clients accept a project, location, and optional endpoint. Ax resolves global, US/EU multi-region, and regional hosts; an explicit base URL takes precedence. Generated packages take a caller-supplied bearer access token and leave ADC refresh to the host application.

GPT-5.6 OpenAI Chat requests can opt into stable explicit prompt-cache breakpoints. Give AxGen a stable promptCacheKey plus contextCache; those forward options reach the provider in every language. Cache reads and writes are normalized separately for usage and catalog-backed cost estimates.

Adaptive balancing

AxBalancer keeps its existing ordered failover behavior by default. Set strategy.type to adaptive to rank equivalent providers per chat request using learned reliability, successful latency, a deadline, and estimated cost. Configure badOutcomeCost in the same currency or unit as the route cost estimate.

Use the native stats-store option for authoritative decision state. The built-in in-memory store can be shared by balancers in one process; multi-process applications can implement AxBalancerStatsStore with an atomic Redis or database update. The routing-event hook is best-effort telemetry, not routing state. Stable route keys are required with a shared store, and namespace plus slice keep unrelated traffic from learning from each other.

Adaptive balancing does not inspect prompt meaning or decide which model is best for a task. The application defines acceptable substitutes through shared logical aliases.

Provider clients

Generated Package Provider Path

The C++ package exposes the AxIR-supported provider surface. Public examples use OpenAI-compatible clients, while internal fixtures cover provider normalization without credentials.

auto client = axllm::OpenAICompatibleClient({
  {"api_key", api_key},
  {"model", "gpt-4.1-mini"},
});

Vertex Gemini

axllm::GoogleGeminiClient client(axllm::object({
    {"api_key", required("GOOGLE_VERTEX_ACCESS_TOKEN")},
    {"project_id", required("GOOGLE_PROJECT_ID")},
    {"region", required("GOOGLE_REGION")},
    {"model", model == nullptr || std::string(model).empty() ? "gemini-3.5-flash" : model},
}));

Use the generated package examples for exact provider API runs, prompt-cached AxGen calls, stream mapping, Responses audio mapping, and realtime event folding for this language.

Embeddings and audio

// Implement embedding calls through the generated AxAI client surface when present.
// Use package conformance coverage to confirm current support for this language.

// Realtime audio over WebSocket — build with -DAXLLM_ENABLE_REALTIME=ON
axllm::OpenAIResponsesClient client(axllm::object({{"model", "gpt-realtime-2"}, {"api_key", std::getenv("OPENAI_APIKEY")}}));
axllm::Value request = axllm::object({{"model", "gpt-realtime-2"}, {"chat_prompt", axllm::array({axllm::object({{"role", "user"}, {"content", "Say hello."}})})}, {"audio", axllm::object({{"output", axllm::object({{"voice", "alloy"}})}})}});
axllm::Value result = client.realtime_chat(request, nullptr);  // one merged turn: transcript + base64 PCM audio
// Realtime models also route transparently through chat(); chat() accepts input_audio parts; transcribe()/speak() do batch STT/TTS.

Practical Notes

• Prefer provider factories over direct provider classes in new code.
• Use model catalog and provider-scoring helpers when choosing between providers.
• Use a multi-service router to dispatch caller-selected model keys; use a balancer for fallback or adaptive operational routing across equivalent services.
• Keep public provider examples separate from internal conformance fixtures.
• Trace provider requests, token usage, estimated cost, and routing decisions in production.

See ai() API.