OpenAI Provider¶

Repository: amplifier-module-provider-openai

GPT model integration for Amplifier via OpenAI's Responses API.

Contract¶

Supported Models¶

• gpt-5.5 — Default model; latest GPT-5 generation
• gpt-5-mini — Smaller, faster GPT-5
• gpt-5-nano — Smallest GPT-5 variant

Configuration¶

[[providers]]
module = "provider-openai"
name = "openai"
config = {
    base_url = null,                       # Optional custom endpoint (null = OpenAI default)
    default_model = "gpt-5.5",
    max_tokens = 4096,
    temperature = 0.7,
    reasoning = null,                      # Reasoning effort: minimal|low|medium|high|xhigh; null = not sent
    reasoning_summary = "detailed",        # Reasoning verbosity: auto|concise|detailed
    truncation = null,                     # null omits the field; OpenAI errors on context overflow.
                                           # Opt in to legacy auto-drop with truncation = "auto" (busts cache).
    prompt_cache_key = "",                 # Stable cache-routing identifier; empty = unset
    prompt_cache_retention = "24h",        # "24h" | "in_memory" | null (use model default)
    enable_response_chaining = "auto",     # "auto" | true | false  (reasoning-model chaining)
    enable_state = false,
    debug = false,                         # Enable standard debug events
    raw_debug = false                      # Enable ultra-verbose raw API I/O logging
}

Note: safety_identifier is intentionally NOT a deployment config field. It is a per-end-user signal (abuse tracking) and must be set per-call via kwargs.

Config Reference¶

Prompt Caching¶

The provider exposes OpenAI's prompt-caching parameters for maximizing cache hit rates.

Default Behavior¶

• prompt_cache_retention = "24h" — Extended GPU-local KV storage on every supported model, instead of OpenAI's per-model "in_memory" default (5–10 min) for gpt-5.4 and below.
• enable_response_chaining = "auto" — For reasoning-capable models, the provider sends store = true and previous_response_id on subsequent turns. Empirical results show 85% prefix cache hit on turn 2 with chaining on, vs 0% off.
• truncation = null — The field is omitted from requests so the cached prefix is never silently rewritten on context overflow. Opt back into legacy auto-drop with truncation = "auto".

prompt_cache_retention Values¶

The capability layer auto-drops values a model would reject (e.g., gpt-5.5+ rejects "in_memory").

enable_response_chaining Values¶

When chaining is active for a reasoning model: store = true is set automatically, previous_response_id is sent on subsequent turns, and encrypted reasoning blocks are not re-inserted inline.

prompt_cache_key¶

OpenAI shards Responses API traffic by hashing the first ~256 input tokens. A stable prompt_cache_key keeps a logical conversation pinned to one machine regardless of small prefix drift.

Deep Research Models¶

The provider supports OpenAI's deep research models which run as background tasks:

Deep research models poll for completion with: - Poll interval: 5.0 seconds - Background timeout: 1800.0 seconds (30 minutes)

Authentication¶

Set your OpenAI API key:

export OPENAI_API_KEY=sk-...

Or for Azure OpenAI:

export AZURE_OPENAI_API_KEY=...
export AZURE_OPENAI_ENDPOINT=https://your-resource.openai.azure.com/

Use provider-azure-openai for managed identity support.

Observability¶

The provider emits standard provider events:

Installation¶

amplifier provider install provider-openai

Or in a bundle:

providers:
  - module: provider-openai
    source: git+https://github.com/microsoft/amplifier-module-provider-openai@main