Vellum

by Vellum AI

Cloud Free Tier paid API Available

Repositioned in 2026 to a "Personal Intelligence" product priced on compute plus credits. Through early 2026 it was an LLM application development platform (Prompt IDE, evaluations, workflows). Vellum was repositioned in 2026 and now presents a product described as "Vellum: Your Personal Intelligence", priced on compute plus pay-as-you-go credits. Through early 2026 it was an LLM application development platform for versioned prompts, evaluations, and deployment, which is retained below as a historical record.

Performance Scores

8.4

1 ranking evaluated

Score range: 8.4 – 8.4

Key Facts

pricing

pricing facts about Vellum
AttributeValueAs ofSource
Vellum pricing (July 2026)As of July 2026, vellum.ai presents a product described as "Vellum: Your Personal Intelligence," priced on compute (vCPU/RAM) plus storage add-ons plus pay-as-you-go credits ($1 = 1 credit): Base (Free, 1 vCPU / 2 GiB + 4 GiB storage), Pro (from ~$50/mo, larger machine), and a self-hosted open-source option. This is a different model from the earlier LLM-application-development platform (Prompt IDE, Evaluations, Workflows) that carried a ~$500/mo Pro tier — the product appears to have changed; verify the current offering at the source.Jul 2026Vellum pricing

General

General facts about Vellum
AttributeValueAs ofSource
StatusRepositioned in 2026: vellum.ai now presents a "Personal Intelligence" product (compute plus credits). It was an LLM-application-development platform (Prompt IDE, evaluations, workflows) through early 2026, retained here for reference.Jul 2026Automation Atlas editorial
PricingAs of July 2026: Base plan free (Small machine 1 vCPU/2 GiB, 4 GiB storage, no credit card); Pro is a $10/month platform fee plus configurable add-ons — compute Medium +$35, Large +$60, XL +$125/mo and storage +$5 to +$120/mo (the page's example Pro configuration totals $50/month); usage credits billed separately pay-as-you-go at $1 = 1 credit on both plans; assistant is open source (MIT) with free self-hosting. Pre-pivot dev-platform pricing (Free $0 / Pro $25/mo / Business $50–$79 per user/mo, as archived Dec 2025–Mar 2026) is retired.Jul 2026Vellum pricing page (live, verified 2026-07-23; historical tiers from web.archive.org snapshots 2025-12-12 and 2026-03-02)
CapabilitiesPersonal AI assistant skills: email triage, calendar, meeting prep, delegation, research, travel booking; memory system and plugin hub (as of July 2026, post-pivot)Jul 2026Vellum homepage
Founded2023Apr 2026Y Combinator
Key DifferentiatorY Combinator-backed; pivoted May 2026 from an LLM dev platform to a personal AI assistant with a layered memory system; $25M raised from Dharmesh Shah, Arash Ferdowsi, Rebel Fund, YCJul 2026Vellum blog — Introducing Vellum (May 7, 2026)

Strengths

  • Built-in evaluation harness with human review and regression testing
  • Production-grade prompt versioning and rollout controls
  • SOC 2 Type II with audit logs and RBAC
  • Multi-model routing across OpenAI, Anthropic, Google, and self-hosted endpoints

Limitations

  • Pricing oriented to mid-market and enterprise — limited free tier
  • Lighter on prebuilt SaaS connectors than agent-first platforms
  • Workflow visual builder is less mature than dedicated agent builders

Based on evaluations in 1 ranking: Best LLM App Platforms for Building AI Agents in 2026

Pricing Plans

Base (Free)

Free

Small machine (1 vCPU, 2 GiB) + 4 GiB storage + PAYG credits

  • Entry compute + storage
  • !Pay-as-you-go credits ($1 = 1 credit)
Get started →
As of Jul 2026 · Source
Most Popular

Pro

$50.00/mo

From ~$50/mo; compute (vCPU/RAM) + storage add-ons + PAYG credits

  • Larger machine, more storage
  • !Credits billed pay-as-you-go
Get started →
As of Jul 2026 · Source

Self-hosted (open source)

Contact us

Open source; you run the infrastructure

  • Self-managed deployment
Get started →
As of Jul 2026 · Source

About Vellum

Vellum was repositioned in 2026 and now presents a product described as "Vellum: Your Personal Intelligence", priced on compute plus pay-as-you-go credits. Through early 2026 it was an LLM application development platform for versioned prompts, evaluations, and deployment, which is retained below as a historical record.

As of 2026, vellum.ai markets an assistant-style product priced on compute (vCPU and RAM) plus storage plus pay-as-you-go credits, with a Base free tier, a Pro tier from about $50/month, and a self-hosted open-source option. This is a different product from the LLM-application-development platform described below. If you are evaluating an LLM-ops or prompt-engineering platform, confirm the current product at the source before relying on either framing (see the Vellum pricing answer for detail).

Historically (founded 2023 in San Francisco, through early 2026), Vellum was an LLM application development platform targeting engineering teams that needed versioned prompts, evaluation pipelines, and deployment infrastructure for shipping AI features into production. That platform combined a Prompt IDE for testing variants across providers (OpenAI, Anthropic, Google, Azure), an Evaluation suite for regression-testing prompt changes against test cases, a Workflows visual builder for chaining LLM calls with code, retrieval, and conditional logic, and a Deployments layer with versioning, monitoring, and request logs. It supported retrieval-augmented generation through managed vector indexes. Its pre-pivot pricing was a free Developer tier, paid Pro plans around $500/month, and custom Enterprise (per public docs, early 2026); as of July 2026 the pivoted personal-assistant product offers a free Base plan and a Pro plan from a $10/month platform fee plus compute and storage add-ons (vendor example configuration about $50/month), with MIT-licensed self-hosting.

Integrations (4)

Anthropic native
Azure OpenAI native
Google Vertex AI native
OpenAI native

Written & reviewed by Rafal Fila · Last updated: · Last verified:

Other AI Agent Platforms Tools

See How It Ranks

Questions About Vellum

What is the best LLM app platform in 2026?

As of April 2026, the leading LLM app platforms are LangChain (most-used Python and JS framework), Vellum (production prompt and eval platform), Langflow (open-source visual builder), Dust (workspace assistants), and LlamaIndex (data-framework for RAG). Choice depends on visual versus code preference and whether teams need eval, RAG, or workspace assistants.

What are the best Langflow alternatives in 2026?

As of April 2026, the leading Langflow alternatives are Flowise (open-source LangChain UI), LangChain itself (code-first SDK), Vellum (production prompt and eval platform), n8n with AI nodes (general workflow plus LLM steps), and CrewAI (multi-agent orchestration in Python). Choice depends on visual versus code preference and whether teams need agents or chains.

What are the best Vellum alternatives in 2026?

As of April 2026, the leading Vellum alternatives are Langflow (open-source visual LangChain builder), LangSmith (LangChain observability and eval), Dust (workspace-grade AI assistants), Relevance AI (low-code AI agents), and Humanloop (prompt management and evaluation). Selection depends on whether teams need prompt eval, agent orchestration, or production deployment.

What are the best Dust alternatives in 2026?

As of April 2026, the leading Dust alternatives are Relevance AI (low-code AI workforce), Lindy (personal AI assistants), Gumloop (visual AI workflows), Relay.app (human-in-the-loop automations), and Glean (enterprise search and assistants). Choice depends on whether teams want internal assistants, agent workforces, or workflow automation.

Learn More