LLM providers¶

AMX talks to LLMs through a single unified interface, so you can swap providers per profile without touching application code or prompts. This page summarises which providers are supported, the trade-offs between them, and where to start when picking the right one for your workload.

Pick a provider¶

Use this short decision tree before reaching for any specific page:

First-time AMX user, prototyping → OpenAI with gpt-4o. The most battle-tested provider; every prompt template and confidence threshold is calibrated against it first.
Cost-sensitive whole-warehouse drafting → Batch mode with gpt-4o-mini or claude-haiku-3-5. ~50% off the live-API rate, async SLA.
Cryptic legacy schemas (transliterated names, abbreviations) → Anthropic with claude-sonnet-4 or extended-thinking on a hard subset.
Big context windows for very wide tables → Gemini with gemini-2.0-flash and column_batch_size: 15.
On-prem / air-gapped → Ollama and local. Llama-3 / Qwen / DeepSeek work; logprob calibration is per-model.

Provider matrix¶

Provider	Default model	Cost lens	Logprobs	Batch API	Key file
OpenAI	`gpt-4o`	Mid (cheap with `mini`)	✓ native	✓ (batch)	`sk-…`
Anthropic	`claude-sonnet-4-20250514`	Mid–High	✓ derived	✓ (batch)	`sk-ant-…`
Gemini	`gemini-2.0-flash`	Low	✓ native	✗ in AMX yet	`AIza…`
OpenRouter	provider/model id	Varies (markup)	varies	✗	`sk-or-…`
DeepSeek	`deepseek-chat`	Very low	✓ native	✗	API key
Ollama / local	`llama3`	Free (compute is yours)	varies	✗	optional

OpenRouter and Kimi are routed through OpenAI-compatible HTTPS — see the wizard prompts; they reuse the OpenAI client under the hood.

Generation defaults that apply across all providers¶

The wizard sets these once per profile (you can edit later in ~/.amx/config.yml):

n_alternatives: 3 — how many candidate descriptions per column. The review wizard offers 1, 2, 3 keys to pick.
column_batch_size: 10 — how many columns AMX packs into one prompt. Bigger = cheaper / column, smaller = higher quality on wide tables.
temperature: 0.2 — deterministic by default. 0.4–0.7 for more variety in alternatives.
logprob_high: 0.85 / logprob_medium: 0.50 — confidence thresholds for the high / medium / low buckets. See /logprob-thresholds.

Per-provider tuning notes live on each provider's page.

Costing rule of thumb¶

For a typical 47-table / 1,283-column schema, drafting descriptions once:

Setup	Approximate cost
Live `gpt-4o-mini`, batch_size 10	$1.00–$1.50
Live `gpt-4o`, batch_size 10	$4.00–$6.00
Batch `gpt-4o-mini`	$0.50–$0.75
Live `claude-sonnet-4`	$5.00–$8.00
Live `gemini-2.0-flash`, batch_size 15	$0.40–$0.80
Local `llama3` on a workstation	$0 (bring your own GPU)

Numbers are illustrative — actual cost depends on column-name length, sample-value length, and the provider's per-token rate at the time. Always run a single-table /run first and check the LLM line for tokens in / out before unleashing it on a warehouse.

Setup walkthroughs¶

Each provider page follows the same template: prerequisites → /add-llm-profile walkthrough with verbatim wizard prompts → sample ~/.amx/config.yml block → verify steps → troubleshooting table → what to read next.

OpenAI — the default; logprob-threshold tuning.
Anthropic — Claude model selection, extended thinking.
Gemini — model picks, safety-filter handling.
Ollama and local — on-prem / air-gapped setup.
Batch mode — async / cheap drafts via OpenAI / Anthropic batch APIs.

What's next¶

Quick start — five-minute install-to-first-comment walkthrough.
Run & Apply — what happens between /run and /apply, including review-wizard keystrokes.
Configuration: env vars — provider-specific env vars (proxies, API base overrides).