Omentir

AI sales agent for automated, personalised LinkedIn outreach

Freemium· Basic $29/mo (1 account, 1 agent, 50 daily leads, 1 campaign) / Startup $59/mo (up to 3 accounts, 3 agents, unlimited leads & campaigns) / Enterprise custom (SSO, dedicated support, unlimited accounts)AgentsClaude Opus 4.5

Visit website →

Best for

Solo founders, small B2B sales teams, and agencies who want an AI agent to run LinkedIn outbound end-to-end and are comfortable connecting it to their own LinkedIn account.

Skip if

Enterprise sales orgs that need multichannel sequencing, deep CRM integration, or a vendor with a long compliance track record; anyone unwilling to accept LinkedIn automation risk.

Omentir is an AI-driven sales automation platform focused on LinkedIn outreach. It plugs into your own LinkedIn account, scores prospects against a written ideal customer profile, then sends connection requests, opening messages, and follow-ups on your behalf while a unified inbox surfaces the replies that actually deserve human attention. Under the hood it uses Claude Opus 4.5 to draft context-aware messages and triage responses, and it exposes an Agent API plus an MCP server so other assistants (Claude, ChatGPT, custom agents) can query leads, kick off campaigns, or read the inbox as tools.

The target user is a solo founder, small B2B sales team, agency, or business-development lead who would otherwise juggle Sales Navigator, a scraping tool, an enrichment provider, a sequencer, and a CRM. Omentir collapses that stack into a single workflow: describe who you want to reach, let the agent build and score a list, review or auto-approve the message drafts, and let it pace sends within LinkedIn safety limits to avoid warnings on the underlying account. Common workflows include cold-outbound campaigns for early-stage SaaS, recruiter sourcing, agency lead-gen on behalf of clients, and using the MCP server to let a broader agent orchestrate outreach as one of several tools. The codebase is MIT-licensed on GitHub, so teams that need to audit prompts, self-host, or extend the scoring logic can do so rather than treating it as a black box.

Editor's take

Omentir is one of the more honest LinkedIn-automation plays we've seen: a frontier model doing the writing, an MCP server so it composes with other agents, and MIT-licensed source you can actually read. The trade-off is scope — it's LinkedIn-only and small — so treat it as a focused outbound agent, not a Sales Navigator replacement.

— The AI Tool Bible editorial team

Pros

✅ Uses a current frontier model (Claude Opus 4.5) for message drafting rather than a cheap fallback
✅ MIT-licensed with full source on GitHub, so prompts and scoring logic are auditable and forkable
✅ Ships an MCP server plus Agent API so other assistants can drive outreach as a tool
✅ Collapses prospecting, sequencing, and inbox triage into a single interface instead of stitching three SaaS tools
✅ Built-in LinkedIn safety limits and human-in-the-loop review reduce account-ban risk
✅ Transparent, low entry price ($29/mo) with no per-lead metering on the Startup tier

Cons

⚠️ LinkedIn-only — no email, X, or multichannel sequencing, so it cannot be a full sales stack on its own
⚠️ Automating from your own LinkedIn account still carries platform-policy risk regardless of safety limits
⚠️ Basic tier is capped at 50 leads/day and one campaign, which most real outbound programs will outgrow quickly
⚠️ Small, newer vendor with limited public deployment history compared to Apollo, Instantly, or HeyReach
⚠️ No native CRM — replies live in Omentir's inbox and must be exported or API-pulled into Salesforce/HubSpot
⚠️ Message quality still depends on the ICP prompt you write; a vague profile will produce generic outreach even with Opus 4.5

Use cases

Automated LinkedIn cold outreachICP-based lead scoringPersonalised connection request draftingFollow-up sequencing on LinkedInUnified reply inbox triageAgency lead-gen on behalf of clientsRecruiter sourcing outreachMCP-tool integration into a broader AI agentFounder-led sales prospecting