Artificial Analysis vs Vellum in 2026

Q: What does Vellum Cloud cost per month?

Vellum is free to download and start . Vellum Cloud subscriptions start at $30 per month for Mighty, $100 per month for Super, or $200 per month for Ultra, with dedicated compute, storage and bundled credits . The Super plan includes a medium computer with 2.5 vCPU, 5 GiB and 30 GB storage .

Artificial Analysis and Vellum compared on what each does, what each costs and what to check before you pick one for tracking or working with AI models.

By SupaKeeper · 2026-10-15

SupaKeeper is published by LunarWerx, which makes some of the tools compared here.

Artificial Analysis provides independent analysis of AI and lets users compare models side by side on intelligence, pricing, speed, latency, benchmarks and more ^[1]. Vellum is a personal assistant that runs locally or in the cloud, works with models from OpenAI, Anthropic, Gemini or open weights via Ollama, and can send emails, manage calendars, write code and run recurring tasks ^[2] ^[2]. Model Odds, made by the team that publishes this site, ranks every Anthropic and OpenAI model family by its estimated chance of a release within 30 days ^[3].

Option	What its sources say it does	Cost	Account or install
Artificial Analysis	Provides independent analysis of AI ^[1]; lets users compare models side by side on intelligence, pricing, speed, latency, benchmarks and more ^[1]	not published ^[1]	not published in the facts ^[1]
Vellum	A personal assistant that can send emails, manage calendars, browse the web, create slides, write code, run deep research, monitor websites and run recurring tasks ^[2]	free to download and start ^[2]; Vellum Cloud starts at $30 per month for Mighty, $100 per month for Super, or $200 per month for Ultra ^[2]	native to Mac and accessible on iOS, Android, web app, voice, email, Telegram, Slack and terminal ^[2]
Model Odds	Ranks every Anthropic and OpenAI model family by its estimated chance of a release within 30 days ^[3]; runs three estimators side by side and publishes the spread as the error bar ^[3]	free, and so is its API, with no account needed ^[3]	a website with a public API ^[3]; a web app that runs in the browser and needs JavaScript ^[3]

The short answer

Artificial Analysis provides independent analysis of AI and lets developers compare models side by side on intelligence, pricing, speed, latency, benchmarks and more ^[1]. Vellum is an assistant that runs locally by default and can send emails, manage calendars, browse the web, create slides, write code, run deep research, monitor websites and run recurring tasks in the background ^[2]. Model Odds ranks every Anthropic and OpenAI model family by its estimated chance of a release within 30 days ^[3]. Pick by what you need: comparing existing models points to Artificial Analysis, running a personal assistant that works across services points to Vellum, and estimating when a model family will release points to Model Odds.

What Artificial Analysis and Vellum are built for

Its page lets users compare models side by side on intelligence, pricing, speed, latency, benchmarks and more ^[1]. Its Artificial Analysis Intelligence Index v4.3.2 incorporates 10 evaluations ^[1]. It maintains an Openness Index assessing how open models are ^[1]. It announced a Cyber Index, and its Cyber Index v1 incorporates 3 evaluations ^[1] ^[1]. Its Coding Agent Index v1.5 incorporates 3 benchmarks ^[1]. It tracks Image Arena and Video Arena leaderboards with 95% confidence intervals ^[1]. It offers a Model Recommender that optimizes recommendations across intelligence, speed and cost ^[1]. It offers Optima, a product for building your own custom benchmark ^[1] ^[1]. Vellum is free to download and start ^[2]. It is open source, so users can inspect its code ^[2]. By default the assistant runs locally, and Vellum Cloud runs each assistant in an isolated container provisioned for that user ^[2]. Conversation history, memory and credentials stay on the user's machine by default and are never used to train models ^[2].

What Model Odds is built for

Model Odds is made by the team that publishes this site. It ranks every Anthropic and OpenAI model family by its estimated chance of a release within 30 days ^[3]. Anthropic and OpenAI each get their own page that lists every past release ^[3]. Each past launch the model uses has to be confirmed by an official post and an independent timeline, and it is listed with its date ^[3]. The estimate is a renewal process fitted on the gaps between recent launches, and it takes into account how long the current wait has already lasted ^[3]. It runs three estimators side by side and publishes the spread between them as the error bar ^[3]. Signals read automatically from company SDKs and docs, and from third-party model catalogues when two of them agree, can only raise the odds ^[3]. The estimator can replay any past date without using data from after that date ^[3]. You can sign up with a confirmed email to get alerts, and every subscriber gets every company's drops at the same time ^[3]. It is a website with a public API ^[3].

Pricing and fees

Artificial Analysis does not state a price in the facts given here, so its cost is not published here. Vellum is free to download and start ^[2], and new users get their first tasks free ^[4]. Vellum Cloud subscriptions start at $30 per month for Mighty, $100 per month for Super, or $200 per month for Ultra, with dedicated compute, storage and bundled credits ^[2]. The Super plan costs $100 per month and includes a medium computer with 2.5 vCPU, 5 GiB and 30 GB storage ^[4]. Model Odds is free, and so is its API, with no account needed ^[3]. If you need the cost of Artificial Analysis for your own use, check its own pages, since the facts here do not state one.

Setting it up and daily use

Vellum is native to Mac and is also accessible on iOS, Android, web app, voice, email, Telegram, Slack and terminal ^[2]. By default the assistant runs locally, and Vellum Cloud runs each assistant in an isolated container provisioned for that user ^[2]. Users can switch between models from OpenAI, Anthropic, Gemini or open weights via Ollama at any time ^[2]. The assistant includes built-in skills for Google Workspace, Slack, Notion and the filesystem, and can be extended with community plugins or custom skills ^[2]. High-stakes actions such as emailing new contacts, modifying important files or spending credits require explicit approval until the user marks them as trusted ^[2]. For Artificial Analysis, the facts describe what its page does but say nothing about installing or setting anything up. Model Odds is a web app that runs in the browser and needs JavaScript ^[3], and it is a website with a public API ^[3]. The website and API share one snapshot that is cached for up to five minutes ^[3]. There is no account needed ^[3].

Things to check before you choose

Ask what you need the tool to do for you. Artificial Analysis compares models side by side on intelligence, pricing, speed, latency, benchmarks and more ^[1], while Vellum is a personal assistant that sends emails, manages calendars, writes code and runs recurring tasks ^[2]. Model Odds estimates when model families will release ^[3]. Ask whether you need a tool that runs on your own machine; Vellum is native to Mac and available on iOS, Android and a web app ^[2], while Model Odds is a web app ^[3]. Check what data leaves your machine; Vellum keeps conversation history, memory and credentials on the user's machine by default ^[2], and Model Odds shares a snapshot cached for up to five minutes ^[3]. Check whether either tool charges for use; Vellum is free to download but Vellum Cloud starts at $30 per month ^[2], and Artificial Analysis does not state a price here.

How to choose

If the need is comparing existing models on intelligence, speed, cost and benchmarks, Artificial Analysis lets users do that ^[1]. If the need is a personal assistant that runs locally and works across services including email, calendars and code, Vellum is built for that ^[2]. If the need is estimating when an Anthropic or OpenAI model family will release within 30 days, Model Odds is built for that ^[3]. If you want verified launch history where each past launch is confirmed by an official post and an independent timeline, Model Odds offers that ^[3]. If you want to switch between models from OpenAI, Anthropic, Gemini or open weights at any time, Vellum lets you do that ^[2]. If you want a free public API that serves per-line odds, launch records and service health, Model Odds offers one ^[3].

Frequently asked questions

Is Model Odds free to use?

Model Odds is free, and so is its API, with no account needed ^[3]. It is a website with a public API, and the API serves per-line odds, launch records and service health ^[3] ^[3]. The facts here give no other price for Model Odds. You can sign up with a confirmed email to get alerts ^[3].

Does Artificial Analysis cost anything?

The facts given here do not state a price for Artificial Analysis. Its page describes independent analysis of AI and comparing models side by side on intelligence, pricing, speed, latency, benchmarks and more ^[1], but says nothing about a fee. Check its own pages if you need the cost.

What does Vellum Cloud cost per month?

Vellum is free to download and start ^[2]. Vellum Cloud subscriptions start at $30 per month for Mighty, $100 per month for Super, or $200 per month for Ultra, with dedicated compute, storage and bundled credits ^[2]. The Super plan includes a medium computer with 2.5 vCPU, 5 GiB and 30 GB storage ^[4].

Can Model Odds estimate when a model family will release?

Yes. Model Odds ranks every Anthropic and OpenAI model family by its estimated chance of a release within 30 days ^[3]. The estimate is a renewal process fitted on the gaps between recent launches, and it takes into account how long the current wait has already lasted ^[3]. It runs three estimators side by side and publishes the spread between them as the error bar ^[3].

Can Vellum run on your own machine without the cloud?

Yes. By default the assistant runs locally, and Vellum Cloud runs each assistant in an isolated container provisioned for that user ^[2]. Conversation history, memory and credentials stay on the user's machine by default and are never used to train models ^[2]. Vellum is open source, so users can inspect its code ^[2].

Model Odds is free, and so is its API, with no account needed. ^[3]

Try Model Odds