Models & Plans

Real numbers. No guesswork. See exactly how many messages, comments, and customers each plan and model can handle.

How Credits Work

Every plan includes a monthly credit allowance. 1 credit = $0.001. When your AI agent replies to a comment or message, it uses credits based on the model and conversation length.

Oro Low
~3
credits per reply
~9,600 input + ~80 output tokens
Oro Mid
~38
credits per reply
Same tokens, higher model cost
Oro High
~60
credits per reply
Maximum intelligence, higher cost

Smart Model Selection: When enabled, your agent automatically picks the cheapest model that fits each message. Simple greetings use Oro Low (~3 credits), complex product questions use Oro Mid or High. This typically saves 60-80% of credits compared to always using Oro High.

Oro Platform Models

Built-in models included with every plan. No API key required.

Oro Low

MOST ECONOMICAL

Fast, efficient replies for greetings, simple questions, and FAQs. Perfect for high-volume pages.

Powered by Gemini 2.5 Flash Lite
Input cost 0.3 credits / 1K tokens
Output cost 1.2 credits / 1K tokens
Avg. per reply ~3 credits

Oro Mid

RECOMMENDED

Balanced quality for most situations. Smart replies, product recommendations, and nuanced conversations.

Powered by Gemini 2.5 Pro
Input cost 3.75 credits / 1K tokens
Output cost 30.0 credits / 1K tokens
Avg. per reply ~38 credits

Oro High

MAXIMUM QUALITY

Complex reasoning, nuanced replies, detailed product explanations. For when quality matters most.

Powered by Gemini 3.1 Pro
Input cost 6.0 credits / 1K tokens
Output cost 36.0 credits / 1K tokens
Avg. per reply ~60 credits

Bring Your Own Key (BYOK)

Use your own API keys for Claude, Gemini, or OpenAI. Available on Starter plan and above. BYOK models use zero Oro credits: you pay the AI provider directly.

Gemini (8 models)

Model Tier Status Best For
Gemini 2.5 FlashRecommendedGABest balance of speed & quality
Gemini 2.5 Flash LiteLiteGACheapest, fastest, high volume
Gemini 2.5 ProProGAComplex reasoning & analysis
Gemini 3 FlashStandardPreviewNext-gen flash model
Gemini 3.1 ProProPreviewLatest & most capable
Gemini 3.1 Flash LiteLitePreviewUltra-cheap next-gen
Gemini 2.0 FlashStandardSunsettingLegacy, use 2.5 instead
Gemini 2.0 Flash LiteLiteSunsettingLegacy, use 2.5 Lite instead

Claude (3 models)

Model Tier Input / MTok Output / MTok
Claude Haiku 4.5Lite$0.80$4.00
Claude Sonnet 4.5Standard$3.00$15.00
Claude Opus 4.6Pro$15.00$75.00

OpenAI (6 models)

Model Tier Input / MTok Output / MTok
GPT-4.1 NanoLite$0.10$0.40
GPT-4o MiniLite$0.15$0.60
GPT-4.1 MiniStandard$0.40$1.60
GPT-4oStandard$2.50$10.00
o4-miniStandard$1.10$4.40
GPT-4.1Pro$2.00$8.00

Real Capacity Per Plan

Based on real production data: average webhook reply = ~9,600 input + ~80 output tokens.

Free

$0/mo
100
credits / month
100
messages / day cap
With Oro Low ~33 replies / month
With Oro Mid ~2 replies / month
With Oro High ~1 reply / month
Customers / day (Low) ~1 customer
Pages / Agents 1 / 1

Best for trying out Oro AI Agent. Credits run out after ~33 Oro Low replies. Use Smart Model Selection to maximize.

Starter

$9/mo
10K
credits / month
500
messages / day cap
With Oro Low ~3,333 replies / month
With Oro Mid ~263 replies / month
With Oro High ~166 replies / month
Customers / day (Low) ~22-37 customers
Pages / Agents 3 / 3

Unlocks BYOK + MCP. With BYOK, credits don't apply, capacity is unlimited (pay provider directly).

Business

$49/mo
100K
credits / month
Unlimited
messages / day
With Oro Low ~33,333 replies / month
With Oro Mid ~2,631 replies / month
With Oro High ~1,666 replies / month
Customers / day (Low) ~222-370 customers
Pages / Agents Unlimited

With Smart Model + caching (6-layer router), handles ~1,000+ customers/day. Ideal for agencies.

Comments vs. Messenger Messages

Comment Replies

Tokens per reply ~9,600 in + ~80 out
Credit cost (Oro Low) ~3 credits
Why higher input? Includes system prompt + knowledge
Typical output Short reply (1-3 sentences)

Messenger Messages

Tokens per reply ~6,400 in + ~45 out
Credit cost (Oro Low) ~2 credits
Why lower input? Shorter conversation context
Avg msgs / customer 3-5 messages per conversation

Customers Per Page Per Day (Oro Low, ~3 msgs each)

Plan Credits Daily Cap Replies / Day
(credit-limited)
Customers / Day
(3 msgs each)
With Caching
(60-70% cached)
Free 100 100 ~1 / day ~1 ~3
Starter $9 10,000 500 ~111 / day ~37 ~92
Pro $19 30,000 2,000 ~333 / day ~111 ~277
Business $49 100,000 Unlimited ~1,111 / day ~370 ~925

6-Layer Smart Router, How We Save Your Credits

Before any message reaches AI, it goes through our smart caching system. Layers 0-3 cost zero credits.

L0
Exact Cache
Identical messages get instant cached reply. Zero tokens. Zero credits.
0 credits
L1
Semantic Cache
Similar-meaning messages match cached replies via vector search (pgvector).
0 credits
L2
Template Engine
Pattern-matched replies for greetings, price inquiries, hours, etc.
0 credits
L3
Intent + Template
AI classifies intent, then matches to a template, minimal AI usage.
0 credits
L4
RAG + Minimal AI
Retrieves relevant knowledge, generates reply with reduced context window.
Reduced
L5
Full AI Passthrough
Complex, unique messages get full AI processing with complete context.
Full cost

Result: In production, 60-70% of messages are handled by layers 0-3 at zero credit cost. This means your plan's capacity effectively multiplies by 2.5-3x. A Pro plan with 30K credits handles what would normally require 75K-90K credits.

Full Feature Comparison

Feature Free Starter $9 Pro $19 Business $49
Pages1310Unlimited
AI Agents1310Unlimited
Messages / Day1005002,000Unlimited
Credits / Month10010,00030,000100,000
Oro Models
BYOK (Claude/OpenAI/Gemini)
MCP Integration
Video Generation
Comment Auto-Reply
Messenger Bot
Smart Model Selection

Start with 100 Free Credits

Download Oro AI Agent and experience AI-powered page management. Upgrade anytime as your business grows.