Dust vs Buzz (Block)
Dust and Buzz (Block) compared on the same dimensions, with Nagent beside them: role, pricing, governance evidence, screenshots and the demonstration to ask for.
Dust is a reusable workspace and orchestration layer at €30 / seat / month; Buzz (Block) is a reusable workspace and orchestration layer at Free and open source (Apache License 2.0). Nagent is multiplayer AI for growth teams: agent teams for marketing, sales and CX in one shared workspace, from $29.99 a month, with autonomy earned on approved work.
At a glance
| Dimension | Nagent ↗ | Dust ↗ | Buzz (Block) ↗ |
|---|---|---|---|
| Stack role | Control plane / Horizontal AI | Control plane / Horizontal AI | Control plane / Horizontal AI |
| Maturity | scaling | scaling | emerging |
| Growth scope | Marketing, Sales, CX | Marketing, Sales, CX | Marketing, Sales, CX |
| Price | From $29.99 per month | €30 / seat / month | Free and open source (Apache License 2.0) |
| Billing basis | Published plans with seats and monthly credits; annual billing saves 33%. Enterprise adds VPC deployment, BYO model and BYO key. | Pro monthly: 8,000 credits. Annual: €24/seat/month equivalent. Excludes VAT; Enterprise custom. | The repository is licensed under the Apache License, Version 2.0, copyright Block, Inc. Self-hosting runs on your own relay infrastructure (the README lists Docker with Postgres, Redis and S3-compatible storage, plus a one-click Railway deploy option); model, hosting and infrastructure costs are separate and set by the providers you choose. No paid plan or hosted-service price was verified in this research pass. |
| Documented statements | Nagent runs a shared workspace for people and agents, Smriti memory that every agent reads, Karma performance records built from each agent's KPIs and governance history, and composable skills, tools, agents and workflows. 17 ready-to-install agents ship today: one super agent, 8 function specialists and 8 industry strategists. Agents run on 19 model providers and 61 models and connect to 998 tools. The marketplace lists 58 published agents in all. 4 patents filed, 3 published. | Dust explicitly markets multiplayer human–agent collaboration. Pro and Max differ in included credits; Max is €150 monthly or €120 on annual billing, with 40,000 credits per seat monthly. Enterprise adds pooled credits, SCIM and audit-log features. | Buzz is a self-hostable workspace built on a Nostr relay, in which every message, reaction, workflow step, review approval and git event is a signed event in one log. Humans and agents each hold a secp256k1 keypair, and agents join channels as members with a Bot role. The README lists as working today the relay, channels, threads, DMs, canvases, media, search, audit log, a Tauri desktop app, an agent-first CLI, an ACP harness for Goose, Codex and Claude Code, YAML workflows with message, reaction, schedule and webhook triggers, and git hosting. Mobile clients and workflow approval gates are listed as being wired up, and SECURITY.md describes Buzz as pre-1.0. |
| Governance evidence | Every action is fired, queued for approval or refused. Approval sits before the irreversible step, destructive changes always ask first, and each agent carries a budget cap and a kill switch. 6 workspace roles (Admin, Sales, Approver, Editor, Writer, Viewer), tenant isolation and an append-only audit trail. Autonomy runs L0 Locked to L4 Fully autonomous and is promoted on approved work, never automatically. SOC 2 Type II and ISO 27001 programmes underway; tenant isolation, RBAC, append-only audit and private VPC available today. | Dust describes separate permissions for agent data access and agent use, plus enterprise retention and deployment options. Validate the selected plan and connector behaviour. | SECURITY.md states that relay connections authenticate via NIP-42, that channel membership is the access-control mechanism with no separate ACL lists, and that events are written to a SHA-256 hash-chained audit log it describes as tamper-evident but not tamper-resistant against an attacker with database write access. The architecture document lists Owner, Admin, Member, Guest and Bot roles, and VISION.md states that workflow approval gates are partially built, with a run that reaches a request_approval step marked Failed (WF-08); VISION.md also states TLS in transit, with at-rest encryption delegated to the storage layer. An open contributor issue (6287, opened 19 August 2026) describes owner confirmation as currently required for agent creation and role edits and proposes configurable trust tiers. Confirm which approval and permission controls work in the release you would deploy, and who operates, secures and backs up the relay. |
| Promoted use cases | Run marketing, sales and CX agent teams with the people who own the number; Give every agent one shared memory and one ledger of decisions; Raise each agent from locked to fully autonomous on approved work, with approval before anything irreversible; Bring in forward deployed engineers and Growth Pods for custom delivery | Research accounts and prepare proposals; Create campaign briefs and performance narratives; Triage support and synthesize company knowledge | Ask the project a question and get an answer with the source threads; Let an agent triage a bug under its own keys, channel memberships and audit trail; Draft release notes from merged pull requests in a workflow, for human review before shipping |
| Review status | Figures from live client campaigns, 2025 to 2026, are published on nagent.ai: Emami Navratna: 24.5M+ reach on a single agent-run campaign; Madmonk AI: 5x more qualified leads; Bigbasket: 70% lower video ad production cost; Sophoz: 3x trial user growth after relaunch. Thirteen brands are named, with four customers quoted by name. These are vendor-published results; ask for a reference in your function and the baseline behind each figure. | G2’s September 2026 seller-invited review by Thibault G. praises model choice, connected knowledge and scheduled automation. A separate 2025 review sample praises reusable agents and support while raising onboarding, documentation and editing concerns. Those historical complaints require re-testing against today’s product. These are attributed experiences, not a representative benchmark or a verified guarantee of results. | No independent review sample was verified in this research pass. Vendor-hosted testimonials are selection-biased customer evidence, not an aggregate rating. |
| Editorial fit | Nagent is the option for teams that want agents and people in one workspace rather than a tool beside the work: shared memory, autonomy that is earned and enforced, published pricing from $29.99 per month, and a delivery team that owns the outcome with you. Evaluate it on the same evidence you would ask of any vendor: shared context, human control and a safe improvement loop. | Dust is a direct horizontal workspace comparator. Its self-service, published seat economics are easier to inspect than a custom managed engagement. Whether that workspace covers a particular end-to-end growth process still depends on configuration, connectors and accountable operators. | Buzz is an open, self-hosted collaboration substrate for people and agents, and its published examples centre on engineering work such as incidents, feature branches and releases. Beside an end-to-end growth team, it sits as infrastructure on which an organisation would configure its own agents, workflows and growth processes. |
| Priority question | Can two colleagues join the same live agent task, change its direction, inspect the memory it used, approve or reject an action and export the audit trail? Nagent demonstrates this in a team room. | Show permission propagation and revocation across a shared conversation, a connected source and a scheduled agent. | In the release we would deploy, can a workflow step pause for human approval before an agent takes a customer-facing action, and who maintains that relay in production? |
| Facts checked | 4 October 2026 | 23 September 2026 | 4 October 2026 |
| Detailed guide | All Nagent comparisons ↗ | Nagent vs Dust ↗ | Nagent vs Buzz (Block) ↗ |
Screenshots and source context
Nagent in detail
Multiplayer AI platform for growth teams: AI agent teams for marketing, sales and CX work with people in one shared workspace, with Smriti memory, Karma scoring and autonomy that is earned, not deployed.
What the vendor documents
Nagent runs a shared workspace for people and agents, Smriti memory that every agent reads, Karma performance records built from each agent's KPIs and governance history, and composable skills, tools, agents and workflows. 17 ready-to-install agents ship today: one super agent, 8 function specialists and 8 industry strategists. Agents run on 19 model providers and 61 models and connect to 998 tools. The marketplace lists 58 published agents in all. 4 patents filed, 3 published.
From $29.99 per month
Published plans with seats and monthly credits; annual billing saves 33%. Enterprise adds VPC deployment, BYO model and BYO key.
Check current commercial source ↗Collaboration and governance
Every action is fired, queued for approval or refused. Approval sits before the irreversible step, destructive changes always ask first, and each agent carries a budget cap and a kill switch. 6 workspace roles (Admin, Sales, Approver, Editor, Writer, Viewer), tenant isolation and an append-only audit trail. Autonomy runs L0 Locked to L4 Fully autonomous and is promoted on approved work, never automatically. SOC 2 Type II and ISO 27001 programmes underway; tenant isolation, RBAC, append-only audit and private VPC available today.
Promoted use cases
- Run marketing, sales and CX agent teams with the people who own the number
- Give every agent one shared memory and one ledger of decisions
- Raise each agent from locked to fully autonomous on approved work, with approval before anything irreversible
- Bring in forward deployed engineers and Growth Pods for custom delivery
Review and customer evidence
Figures from live client campaigns, 2025 to 2026, are published on nagent.ai: Emami Navratna: 24.5M+ reach on a single agent-run campaign; Madmonk AI: 5x more qualified leads; Bigbasket: 70% lower video ad production cost; Sophoz: 3x trial user growth after relaunch. Thirteen brands are named, with four customers quoted by name. These are vendor-published results; ask for a reference in your function and the baseline behind each figure.
Editorial fit
Nagent is the option for teams that want agents and people in one workspace rather than a tool beside the work: shared memory, autonomy that is earned and enforced, published pricing from $29.99 per month, and a delivery team that owns the outcome with you. Evaluate it on the same evidence you would ask of any vendor: shared context, human control and a safe improvement loop.
Priority question for the demo: Can two colleagues join the same live agent task, change its direction, inspect the memory it used, approve or reject an action and export the audit trail? Nagent demonstrates this in a team room.
Dust in detail
Collaborative enterprise AI workspace for teams to build and use connected agents.
What the vendor documents
Dust explicitly markets multiplayer human–agent collaboration. Pro and Max differ in included credits; Max is €150 monthly or €120 on annual billing, with 40,000 credits per seat monthly. Enterprise adds pooled credits, SCIM and audit-log features.
€30 / seat / month
Pro monthly: 8,000 credits. Annual: €24/seat/month equivalent. Excludes VAT; Enterprise custom.
Check current commercial source ↗Collaboration and governance
Dust describes separate permissions for agent data access and agent use, plus enterprise retention and deployment options. Validate the selected plan and connector behaviour.
Promoted use cases
- Research accounts and prepare proposals
- Create campaign briefs and performance narratives
- Triage support and synthesize company knowledge
Review and customer evidence
G2’s September 2026 seller-invited review by Thibault G. praises model choice, connected knowledge and scheduled automation. A separate 2025 review sample praises reusable agents and support while raising onboarding, documentation and editing concerns. Those historical complaints require re-testing against today’s product. These are attributed experiences, not a representative benchmark or a verified guarantee of results.
Editorial fit
Dust is a direct horizontal workspace comparator. Its self-service, published seat economics are easier to inspect than a custom managed engagement. Whether that workspace covers a particular end-to-end growth process still depends on configuration, connectors and accountable operators.
Priority question for the demo: Show permission propagation and revocation across a shared conversation, a connected source and a scheduled agent.
Buzz (Block) in detail
Open-source, self-hostable workspace from Block where people and AI agents share channels on a Nostr relay.
What the vendor documents
Buzz is a self-hostable workspace built on a Nostr relay, in which every message, reaction, workflow step, review approval and git event is a signed event in one log. Humans and agents each hold a secp256k1 keypair, and agents join channels as members with a Bot role. The README lists as working today the relay, channels, threads, DMs, canvases, media, search, audit log, a Tauri desktop app, an agent-first CLI, an ACP harness for Goose, Codex and Claude Code, YAML workflows with message, reaction, schedule and webhook triggers, and git hosting. Mobile clients and workflow approval gates are listed as being wired up, and SECURITY.md describes Buzz as pre-1.0.
Free and open source (Apache License 2.0)
The repository is licensed under the Apache License, Version 2.0, copyright Block, Inc. Self-hosting runs on your own relay infrastructure (the README lists Docker with Postgres, Redis and S3-compatible storage, plus a one-click Railway deploy option); model, hosting and infrastructure costs are separate and set by the providers you choose. No paid plan or hosted-service price was verified in this research pass.
Check current commercial source ↗Collaboration and governance
SECURITY.md states that relay connections authenticate via NIP-42, that channel membership is the access-control mechanism with no separate ACL lists, and that events are written to a SHA-256 hash-chained audit log it describes as tamper-evident but not tamper-resistant against an attacker with database write access. The architecture document lists Owner, Admin, Member, Guest and Bot roles, and VISION.md states that workflow approval gates are partially built, with a run that reaches a request_approval step marked Failed (WF-08); VISION.md also states TLS in transit, with at-rest encryption delegated to the storage layer. An open contributor issue (6287, opened 19 August 2026) describes owner confirmation as currently required for agent creation and role edits and proposes configurable trust tiers. Confirm which approval and permission controls work in the release you would deploy, and who operates, secures and backs up the relay.
Promoted use cases
- Ask the project a question and get an answer with the source threads
- Let an agent triage a bug under its own keys, channel memberships and audit trail
- Draft release notes from merged pull requests in a workflow, for human review before shipping
Review and customer evidence
No independent review sample was verified in this research pass. Vendor-hosted testimonials are selection-biased customer evidence, not an aggregate rating.
Editorial fit
Buzz is an open, self-hosted collaboration substrate for people and agents, and its published examples centre on engineering work such as incidents, feature branches and releases. Beside an end-to-end growth team, it sits as infrastructure on which an organisation would configure its own agents, workflows and growth processes.
Priority question for the demo: In the release we would deploy, can a workflow step pause for human approval before an agent takes a customer-facing action, and who maintains that relay in production?
How to run the evaluation
Run the same bounded pilot on Dust and Buzz (Block) and on Nagent, with the inputs, the required output and the measure fixed before the first demo. Ask each vendor to show who can propose, approve, execute, interrupt and audit the action, and keep the evidence.
Account research to approved follow-up
Give the team an account record, a current product brief and a small set of approved customer references. Ask it to identify a relevant problem, distinguish observed facts from inferred needs, and prepare a follow-up for review. Then change one source and require a second person to correct the draft without restarting the research.
Measure: source accuracy, reviewer minutes, approved follow-ups and CRM consistency.
Campaign brief to coordinated delivery
Start with one campaign objective, audience, budget constraint and brand guide. Require research, a brief, assets and a release checklist with distinct human owners. Introduce an unresolved claim midway through the process. The system should retain useful work while preventing that claim from silently propagating into every asset.
Measure: accepted assets, revision burden, approval latency and missing handoffs.
Customer insight to shared learning
Provide anonymized support themes and sales objections. Ask for a synthesized insight, a proposed campaign adjustment and a knowledge update. A marketing owner and a CX owner should review the same underlying evidence. The exercise tests cross-functional context and accountability, not merely summarization quality.
Measure: evidence coverage, disagreement resolution and reuse of approved learning.
The evaluation owner should be a growth operations lead. Use accepted work completed across tools as the business target.
Frequently asked questions
Dust vs Buzz (Block): what is the difference?
Dust is a reusable workspace and orchestration layer; Collaborative enterprise AI workspace for teams to build and use connected agents. Buzz (Block) is a reusable workspace and orchestration layer; Open-source, self-hostable workspace from Block where people and AI agents share channels on a Nostr relay. Both sit in the same stack role, so delivery scope, commercial basis and governance decide.
How do Dust and Buzz (Block) compare on price?
Dust: €30 / seat / month. Pro monthly: 8,000 credits. Annual: €24/seat/month equivalent. Excludes VAT; Enterprise custom. Buzz (Block): Free and open source (Apache License 2.0). The repository is licensed under the Apache License, Version 2.0, copyright Block, Inc. Self-hosting runs on your own relay infrastructure (the README lists Docker with Postgres, Redis and S3-compatible storage, plus a one-click Railway deploy option); model, hosting and infrastructure costs are separate and set by the providers you choose. No paid plan or hosted-service price was verified in this research pass. Nagent publishes 4 plans, from $29.99 a month, with monthly credits as the unit of agent work.
Which has stronger governance: Dust or Buzz (Block)?
Dust: Dust describes separate permissions for agent data access and agent use, plus enterprise retention and deployment options. Validate the selected plan and connector behaviour. Buzz (Block): SECURITY.md states that relay connections authenticate via NIP-42, that channel membership is the access-control mechanism with no separate ACL lists, and that events are written to a SHA-256 hash-chained audit log it describes as tamper-evident but not tamper-resistant against an attacker with database write access. The architecture document lists Owner, Admin, Member, Guest and Bot roles, and VISION.md states that workflow approval gates are partially built, with a run that reaches a request_approval step marked Failed (WF-08); VISION.md also states TLS in transit, with at-rest encryption delegated to the storage layer. An open contributor issue (6287, opened 19 August 2026) describes owner confirmation as currently required for agent creation and role edits and proposes configurable trust tiers. Confirm which approval and permission controls work in the release you would deploy, and who operates, secures and backs up the relay. Nagent puts approval before any irreversible action, gives each agent a budget cap and a kill switch, and keeps an append-only audit trail.
Is Nagent an alternative to Dust or Buzz (Block)?
Nagent is a multiplayer AI platform where agent teams for marketing, sales and CX work with people in one shared workspace. It overlaps with Dust and Buzz (Block) where a growth team wants agents to do the work, and differs in operating model, governance and delivery; many teams run it beside a a reusable workspace and orchestration layer.
What should I ask Dust and Buzz (Block) in a demo?
Dust: Show permission propagation and revocation across a shared conversation, a connected source and a scheduled agent. Buzz (Block): In the release we would deploy, can a workflow step pause for human approval before an agent takes a customer-facing action, and who maintains that relay in production? For Nagent, ask to see a team room with several people, an approval taken in the Inbox and the audit entry it leaves.
Sources and research notes
- Nagent · product ↗
- Nagent · commercial / feature reference ↗
- Nagent · platform, workspace and security ↗
- Nagent · customer results ↗
- Dust · product ↗
- Dust · commercial / feature reference ↗
- Pods overview · shared conversations, tasks and files ↗
- Pod conversations · participation and search ↗
- Independent reviews · G2, current reviewer index ↗
- Independent reviews · G2, historical pros and cons ↗
- Buzz (Block) · product ↗
- Buzz (Block) · commercial / feature reference ↗
- Buzz VISION.md · GitHub, Block ↗
- Buzz SECURITY.md · GitHub, Block ↗
- Buzz ARCHITECTURE.md · GitHub raw file, Block ↗
- Issue 6287, configurable agent approval and trust tiers · GitHub ↗
- Nagent · security and compliance ↗