MatterAI: Pricing, Features & Best Use Cases

Our take: Choose MatterAI if you want a single OpenAI-compatible API for open-source and third-party models with a stated 30 percent discount and a model catalog that lists per-million-token rates.

Main limitation: The pricing page was empty and docs and privacy pages returned errors, so plan tiers, retention and training terms could not be confirmed.

By the AI Tool Atlas editors. Based on the vendor's official documentation; we have not tested this product hands-on. What “documented overview” means

Is MatterAI right for you?

Choose MatterAI if you want a single OpenAI-compatible API for open-source and third-party models with a stated 30 percent discount and a model catalog that lists per-million-token rates. Based on the vendor's documentation, not on a hands-on test.

Best for: Developers wanting a single API for several models

Also a good fit for

Teams wanting automated code reviews

Consider another tool if plan prices on the pricing page are not visible without JavaScript rendering

Or if

Token-priced API usage is metered
Account sign-in required for API keys

MatterAI pricing and usage limits

Prices checked Oct 2, 2026 on the official US pricing page. Official pricing ↗ Prices and limits change often.

The official MatterAI pricing page as we saw it. matterai.so, captured Oct 2, 2026

Free access and trial conditions

Gemma 4 is listed as a Free Plan model in the catalog.

Paid plans: what changes as you upgrade

API pay as you go api

Usage-based source (opens in a new tab)

Pricing page was empty; plan tiers not confirmed

For: Developers calling hosted models

Limits

30 percent discount on tokens stated on homepage

Includes

OpenAI-compatible chat completions
Automatic prompt caching discounts up to 90 percent

API included

Usage caps, credits and extra costs

API usage is billed per million tokens. First purchase offer: pay $5 and get $20 in API credits.

What is MatterAI, and how does it work?

MatterAI, from Gravity Cloud Services, Inc., offers hosted inference for open-source and third-party models through one OpenAI-compatible API, with a stated 30% discount on listed token prices. Its coding products include the Orbital AI coding agent and IDE, the OrbCode terminal CLI, an MCP server, and AI code reviews for pull requests.

MatterAI's official site: how the product presents itself. matterai.so, captured Oct 2, 2026

What you provide and what you get back

What you give MatterAI and what comes back, from its documentation

You provide

Chat completion requests to the API
Prompts typed in the OrbCode terminal CLI
Code and pull requests for review

You get back

Model completions
Code edits and commands with approval
Automated pull request reviews

Where you can use it

Platforms: Web, API
Account: Required
Coding tools: OrbCode CLI, Orbital IDE, AI code reviews
Terminal install: npm i -g @matterailab/orbcode
Listed models: Gemma 4, Muse Spark 1.3, DeepSeek V4 Flash, GLM 5.3, GPT-5.6, Gemini 3.8 Flash
Infrastructure partners: Baseten, Fireworks AI, Cloudflare AI
OrbCode license: MIT

Where MatterAI is most useful

Scenarios are worked out by our editors from the documentation; the actual result depends on your input and plan.

Calling open-source and third-party LLMs through one API

Who: Developers building LLM apps
You provide: Requests to the chat completions endpoint using a MatterAI API key
Expected result: Completions from catalog models with per-token billing
Plan needed: Pay as you go API
Limitation: Some catalog models are labeled Contributor or Free Plan source (opens in a new tab)

Agentic coding in the terminal

Who: Developers who work in a terminal
You provide: Natural-language coding tasks in OrbCode
Expected result: File edits and commands run after your approval
Plan needed: Not stated
Limitation: OrbCode plan requirements are not confirmed source (opens in a new tab)

AI coding assistance in an IDE

Who: Engineering teams reviewing code
You provide: Pull requests
Expected result: Automated review comments and fix commits
Plan needed: Not stated
Limitation: Review plan limits are not published source (opens in a new tab)

MatterAI features and their limits

Feature overview on the official site. matterai.so, captured Oct 2, 2026

OpenAI-compatible chat completions endpoint

Accepts standard chat completion calls at a MatterAI base URL so existing OpenAI-style clients can point to it. Reduces code changes when switching providers.

Access: API key requiredLimit: Only chat completions is described on the homepage. source (opens in a new tab)

Catalog of models with per-million-token pricing

Lists models with provider, context length and prompt, completion and cache-read prices per million tokens after a 30 percent discount. Makes cost comparison possible before calling a model.

Access: Public on the homepageLimit: Prices shown are illustrative of the date fetched and may change. source (opens in a new tab)

Free-plan Gemma 4 model listed

Includes Gemma 4, shown at free prompt, completion and cache rates. Allows testing without paying per token.

Access: Labeled Free PlanLimit: Free plan limits are not documented. source (opens in a new tab)

OrbCode CLI

Provides MIT-licensed agentic coding in the terminal with streaming TUI, edit and command approvals and headless mode. Brings agentic coding to terminal workflows.

Access: Installed with npmLimit: Plan or credit needs are not confirmed. source (opens in a new tab)

MatterAI: strengths and limits

Reasons to choose MatterAI

Model catalog lists per-1M-token prices publicly on the homepage Reduces the effort of comparing model costs. documented source (opens in a new tab)
OpenAI-compatible API endpoint Reduces migration effort for apps already using an OpenAI-style client. documented source (opens in a new tab)
Free-plan model listed in the catalog Lets developers test calls without token charges. documented source (opens in a new tab)
Documentation and changelog are published Gives developers somewhere to look for details, although the docs URL failed when fetched. documented source (opens in a new tab)

Check before paying for MatterAI

Plan prices on the pricing page are not visible without JavaScript rendering Users must contact or sign in to learn actual plan tiers. documented source (opens in a new tab)
Token-priced API usage is metered Costs scale with usage and can be hard to forecast. documented source (opens in a new tab)
Account sign-in required for API keys Developers must create an account before testing the API. documented source (opens in a new tab)
Some models are labeled contributor or tiered rather than free Cost-sensitive users must check each model label before relying on it. documented source (opens in a new tab)

How to use MatterAI for calling hosted models through one API

Set up MatterAI

Open matterai.so and choose Get API Key.
Sign in to create the key.
Copy a model ID from the catalog.
Set your client base URL to https://api2.matterai.so/v1.

Example: calling hosted models through one API

Send a chat completion request using the key and the gemma-4 model ID.
Read the response and token usage.
Switch the model ID to a paid catalog model and repeat.

Review and export the result

Confirm the request succeeds with a standard OpenAI-style client.
Compare billed tokens with the catalog price per 1M tokens.

Example prompt suggested by our editors (not a recorded output of the product)

Summarize this changelog in five bullet points.

MatterAI API

Plan comparison on the official pricing page. matterai.so, captured Oct 2, 2026

API access and separate billing

MatterAI has no public API in its official documentation. Homepage lists base URL https://api2.matterai.so/v1; docs page returned 404 when fetched. API documentation ↗

Other tools to consider instead of MatterAI

"Better suited" is based on documented features and pricing, not on a head-to-head test.

Cursor: better suited to

GitHub Copilot: better suited to

Devin Desktop (formerly Windsurf): better suited to

Alternatives
Alternative	Pricing	Limitation
Cursor	—	—	Compare
GitHub Copilot	—	—	Compare
Devin Desktop (formerly Windsurf)	—	—	Compare

Explore MatterAI further

Best MatterAI alternativesTools for the same job, by need and starting price.Best AI coding toolsOur picks in this category, compared by price and free tiers.

Tools and guides near MatterAI

More coding tools

Guides and comparisons

MatterAI user reviews

No user reviews yet.

Used MatterAI? Be the first to share your experience.

Write a review

MatterAI FAQ

Is the MatterAI API compatible with OpenAI clients?

Yes, the homepage describes one OpenAI-compatible API with the base URL https://api2.matterai.so/v1. source (opens in a new tab)

What discount does MatterAI apply to model tokens?

The homepage states a 30 percent discount below upstream list prices, shown as strikethrough prices in the catalog. source (opens in a new tab)

Which MatterAI model is listed as free?

Gemma 4 from Google, labeled Free Plan with a 262k context and free prompt, completion and cache rates. source (opens in a new tab)

What is MatterAI first purchase offer?

The homepage says to pay $5 on a first purchase and receive $20 in API credits. source (opens in a new tab)

What is OrbCode from MatterAI?

OrbCode is an MIT-licensed terminal CLI for agentic coding, installed with npm i -g @matterailab/orbcode. source (opens in a new tab)

Sources and review methodology

Official sources and verification dates

MatterAI homepage homepage · checked Oct 2, 2026
MatterAI pricing homepage · checked Oct 2, 2026
MatterAI documentation homepage · checked Oct 2, 2026

Not verified: Plan tiers and prices (pricing page empty); Privacy, retention and training terms (privacy page 404); Documentation content (docs page 404); Orbital IDE platforms.

How we documented and maintain this page

Every fact about MatterAI above comes from the 3 official pages listed here, last compared on Oct 2, 2026; we have not run MatterAI ourselves, so it has no editorial score. How we check and re-check pages.

Article updated Oct 2, 2026; prices checked Oct 2, 2026. Outbound links to vendors are plain links; we do not earn commissions. How we review · Report an error