Get started
Sovereign Compute & Fleet Quotas

Predictable Pricing for Autonomous Fleets

Free while it is just you. Per seat when a second operator joins the verification loop. Custom when an enterprise requires centralized governance ceilings, air-gapped models, and self-hosted microVMs.

Interactive Fleet Estimator

Estimate Your Compute & Fiber Scale

65% compute savings with Jev Substrate
Concurrent Agent Fibers: 8 Fibers
1 Fiber 8 Team 16 Scale 32 Fleet
Monthly Token Budget: 2.5M Tokens/mo
200k 2.5M 10M 25M+
💡 How Jev saves compute: Helix uses the local Jev decision substrate for sub-second classification, memory recall, and loop supervision. Only high-uncertainty turns route to frontier cloud models, reducing API costs by up to 70%.
Estimated Monthly Cloud vs Helix Cost Curve
Merchant of Record: Paddle Billing (sandbox)

Personal

One operator, your own machine or a single Docker host.

Free forever

No card. One seat, one workspace.

Includes
  • Unlimited sessions on your own hardware
  • Local, Docker, and WASM sandboxes
  • Full event log with replay, branching, and Workflow Canvas
  • Cross-platform Helix CLI, Local Cockpit & DASP 1.0 SDKs
  • Dynamic Cognitive Tiering (✨ Auto) & encrypted BYOK across 19 providers
  • Approvals in the app
  • One domain pack
  • 7-day event retention

Team

Most teams

A working group that shares approvals, artifacts and an audit trail.

$24 per seat / month

Billed monthly via Paddle. Two-seat minimum.

Everything in Personal, plus
  • Shared workspaces and saved views
  • Firecracker microVM sandboxes with hibernation & Browser CDP
  • Approvals inbox with SLA and auto-deny
  • Governance ceiling and per-run spend limits
  • 4 Core domain packs (Engineering, Sales, Marketing, Science)
  • Prepaid Helix AI credit packs ($10, $25, $100) + BYOK
  • Synapse Epistemic Core with openCypher & JTMS truth maintenance
  • Beacon Cross-Harness skill synthesis & 90+ MCP tools
  • 90-day event retention
  • Slack approvals and daily digests
  • Workspace-owned automations

Enterprise

Multiple workspaces under one governance ceiling you set centrally.

$299 base / month

Custom volume pricing, invoicing, procurement review, or self-serve base tier.

Everything in Team, plus
  • Org-level governance ceiling no workspace can exceed
  • All 7 domain packs including Regulated Packs (Healthcare, Finance, Legal) + custom Hex packs
  • Synapse EDW connectors (BigQuery, Snowflake, Databricks)
  • SSO with SAML and SCIM provisioning
  • Bring your own model provider, private VPC, or air-gapped local substrate
  • Self-hosted or private-region deployment
  • Signed audit export, GDPR/DPDP erasure receipts, and configurable retention
  • DPA, subprocessor review, security questionnaire
  • Named support with response commitments

Prices in USD, exclusive of tax. Every workspace starts with $1.00 in free Helix AI credits. Bring your own encrypted API key (BYOK) across 19 cloud and local providers with no additional platform fees. Helix is a product of Codemint Labs LLP. Orders are processed by our Merchant of Record, Paddle.com, subject to our Terms of Service, 14-Day Refund Policy, and Privacy Policy.

19 Integrated Model Providers · Dynamic Cognitive Tiering (✨ Auto)

Unified AI Routing & Bring-Your-Own-Key Flexibility

Route turns automatically with ✨ Auto (< 15ms Jev complexity classification and resilient tier fallback) using prepaid Helix Managed Credits with automatic prompt-cache optimization, or connect your own AES-256-GCM encrypted BYOK credentials across 19 cloud and local providers with no additional platform fees. Deterministic cron jobs, triggers, and heartbeats that do not invoke a model cost $0.00.

$1.00 Free Starter Credit Direct Encrypted BYOK
Cognitive Tier Representative Models Workload & Routing Profile Helix Managed Credits BYOK / Local Substrate
⚡ Fast Tier Gemini 2.0 Flash, GPT-4o mini, Claude 3.5 Haiku, Llama 3.3 70B Sub-second triage, classification & summaries Included (Prompt-cache optimized) Direct provider billing
⚖️ Standard Tier Claude 3.5 Sonnet, GPT-4o, GPT-4.1, DeepSeek V3, Command R+ Daily coding, tool orchestration & synthesis Included (Prompt-cache optimized) Direct provider billing
🧠 Frontier Tier Claude 3.7 Sonnet, Claude Opus 4, OpenAI o3 / o1, Gemini 2.5 Pro, DeepSeek R1, Grok 3 Deep multi-file architecture, proofs & planning Included (Resilient fallback) Direct provider billing
🏠 Local / Private Ollama, LM Studio, vLLM (Self-hosted GPU clusters) Air-gapped & zero-egress offline execution Local hardware (Zero cloud egress) 100% On-Premise

Detailed Plan Comparison

Capability Personal Team Enterprise
Sessions, workflows and replay
Supervised planner and worker fibers
Event log replay from event 0
Trajectory branching and Counterfactual Studio
Unified chat & visual DAG Workflow Canvas
Cross-platform CLI & 14-module Local Cockpit (:4005)
Event retention 7 days 90 days Configurable
Concurrent worker fibers 3 16 Negotiated
Synapse & Cognitive Routing
Dynamic Cognitive Tiering (✨ Auto) across 19 providers
Resilient tier fallback & offline local model fallback
Synapse Epistemic Context Graph —
Bitemporal time coordinates (Valid/System) & JTMS —
openCypher query engine & W3C PROV-O lineages —
Enterprise EDW connectors (BigQuery/Snowflake/Databricks) — —
Domain Packs, SDKs & Cross-Harness
Domain Packs supported 1 Core Pack 4 Core Packs All 7 Packs (incl. Regulated)
HIPAA/FHIR Healthcare, SOX Finance & Legal Packs — —
DASP 1.0 SDKs (Elixir, TypeScript, Python) & 90+ MCP tools
Beacon trace ingestion (Claude Code, Cursor, Cline) —
Jev System-One logit decisions (<500ms) —
Sandboxing & Execution Worlds
Local process isolation & Docker containers
WASM fuel-metered plugin runtime
Firecracker / Sprites microVMs with hibernation —
Browser CDP Screencast & Computer Use —
Self-hosted sandbox hosts & Cluster Mesh — —
Security, Privacy & Governance
Eight-gate zero-trust tool pipeline
AES-256-GCM encrypted BYOK credential vault
Self-serve GDPR/DPDP workspace export & erasure receipts
Approvals inbox with SLA and auto-deny —
Workspace governance ceiling (Policy ∩ Profile) —
Org ceiling & cognitive tier policy caps — —
Signed audit export & SIEM streaming — —
Team and access
Seats 1 Unlimited Unlimited
Shared workspaces —
Roles: owner, admin, member, viewer —
SSO (SAML) and SCIM — —
Support
Community
Email support —
Named contact with response SLA — —

Frequently Asked Questions

What counts as a seat?

Anyone who can open a session or decide an approval in a workspace. Viewers who only read artifacts and audit entries are free on Team.

What happens when a run exceeds its spend ceiling?

The run stops and files an approval request for more headroom rather than continuing quietly. The ceiling is per run, and a workspace owner decides.

Do I need my own model API key?

No — every workspace starts with $1.00 in free Helix AI credits routed across Anthropic, OpenAI, Google Gemini, DeepSeek, Groq, xAI, and Mistral through the unified Helix AI Router (with automatic prompt-cache optimization and resilient fallback). You can top up $10, $25, or $100 credit packs anytime, or connect your own AES-256-GCM encrypted API keys (BYOK) across all 19 cloud and local providers with no additional platform fees.

Can I self-host?

Enterprise deploys into your own infrastructure or a private region. The sandbox, model, memory, storage and notification seams each bind to providers you control.

What happens to my data if I cancel?

Sessions, artifacts and the event log are exportable anytime via self-serve GDPR/DPDP workspace export and remain accessible for 30 days after cancellation. You can also trigger cascading right-to-erasure with a cryptographic SHA-256 receipt.

How do billing, taxes, and refunds work?

Our order process is conducted by our online reseller Paddle.com, which acts as the Merchant of Record for all Helix orders and handles global payments, localized currency pricing, and Sales Tax / VAT / GST invoicing. Initial subscription purchases and unused prepaid credit packs are covered by a 14-day refund policy in accordance with Paddle's Buyer Terms. Subscriptions can be upgraded, downgraded, or canceled anytime from Settings → Governance.

Is there a discount for annual billing?

Team is two months cheaper paid annually. Enterprise is annual by default with volume pricing.

Ready to deploy your agent fleet?

Start on Personal and keep full ownership of your event log and artifacts. Moving to Team is seamless at any billing cycle.