# Aissist.io — Multi-agent AI that resolves support and sales conversations inside your existing helpdesk.

> Source: The Agents Index — https://theagentsindex.com/aissist (structured, researched, re-verified)
> Facts last verified: 2026-10-24

Aissist.io is an AI layer that sits on the helpdesk a team already runs and resolves support and sales conversations end to end with coordinating specialist agents. It reads connected knowledge and systems, acts in them, and hands difficult cases to a person.

| Fact | Value |
| --- | --- |
| Website | https://aissist.io |
| Price | free, then from $0.09/interaction pay-as-you-go |
| Free tier | Yes |

Free plan: up to 1,000 tickets per month, every month, no credit card required; runs on the 'free' engine rather than the advanced engine used on Growth.

## What it does

**Done-for-you service** · No — You run the software yourself: the vendor states setup is self-serve in roughly ten minutes with no mandatory implementation contract, and only the Enterprise tier adds dedicated AI engineer support.

> Pricing is public (you're reading it), setup is self-serve in about 10 minutes, and there's no mandatory implementation contract or sales gate to get started.
> — https://aissist.io/product/aissist-straight-answers · Section 'No. 02 — The ledger', answer to "It's hidden-pricing enterprise software with a months-long implementation"

## How much it does unattended

**Runs autonomously** · Yes — The vendor positions it as completing cases end to end rather than drafting, with resolution counted only when no human escalation happened, and it judges by itself when to hand over.

> A resolution is counted when the AI fully resolves the request without escalation after at least one interaction of genuine user inquiry (chitchat does not count).
> — https://aissist.io/pricing · Pricing FAQ, 'What counts as a resolution?'

**Multi-agent** · Yes — Specialised sub-agents each own a domain such as policy, product knowledge or escalation logic and collaborate on the same case on the vendor's own Multi-Agent Platform.

> Specialized sub-agents each own a domain — policy, product knowledge, troubleshooting, system actions, escalation logic — and collaborate on the same case.
> — https://aissist.io/product/agentmesh-agentic-ai-workforce · Section 'What makes AgentMesh different under the hood?'

**Agent permissions** · Partly — It acts freely inside customer-set guardrails and escalates sensitive or out-of-policy cases, and optimisation changes touching policy, brand or compliance are held for human sign-off rather than applied automatically.

> Ship with approval What works is rolled out. Changes to policy, brand or compliance wait for your sign-off.
> — https://aissist.io/ · Homepage, 'Self-evolve AI' section, step 03

## Which models it runs on

**Claude** · Yes — The vendor's own lock-in audit of itself names Claude among the models behind the product, while noting customers cannot bring their own model.

> Moderate — FAQ names OpenAI, Grok, Claude, Gemini; no BYO-model, and the trust center lists only OpenAI
> — https://aissist.io/insights/ai-customer-service-vendor-lock-in · Vendor scoring table, Aissist.io row, 'Model dependency' cell

**GPT** · Yes — The same self-assessment names OpenAI among the underlying models and says the trust centre lists only OpenAI, so GPT models are in use under the hood.

> Moderate — FAQ names OpenAI, Grok, Claude, Gemini; no BYO-model, and the trust center lists only OpenAI
> — https://aissist.io/insights/ai-customer-service-vendor-lock-in · Vendor scoring table, Aissist.io row, 'Model dependency' cell

**Model choice** · No — The vendor's own audit records that there is no bring-your-own-model path, so the model mix is the vendor's choice rather than the buyer's.

> Moderate — FAQ names OpenAI, Grok, Claude, Gemini; no BYO-model, and the trust center lists only OpenAI
> — https://aissist.io/insights/ai-customer-service-vendor-lock-in · Vendor scoring table, Aissist.io row, 'Model dependency' cell

## Where you use it

**In a browser** · Yes — Buyers sign up and work in a hosted console, and the Pulse insight product renders dashboards and an ask interface in the browser.

> Ask Engine Just ask — it identifies the issue and the root cause in real time. AI performance Resolution, CSAT and cost, live.
> — https://aissist.io/ · Homepage 'Reveals' section, Pulse dashboard captions

## Whose machine it runs on

**Open source** · No — The engine is described as the vendor's proprietary Multi-Agent Platform, and no repository or licence is published anywhere on the site.

> AgentMesh runs on Aissist.io's proprietary Multi-Agent Platform, or M.A.P.
> — https://aissist.io/product/agentmesh-agentic-ai-workforce · Section 'What makes AgentMesh different under the hood?', first paragraph

## What it costs to run

**How it meters** · Yes — Billing is metered per AI interaction with a per-resolution cap and the buyer pays whichever is lower, charged after usage rather than prepaid, and handed-over cases are not charged.

> Charging $0.09 per interaction, capped at $0.60 per resolution — you pay the lower of the two — keeps it fair everywhere, and billing is always post-usage, never pre-paid.
> — https://aissist.io/pricing · Pricing FAQ, 'Why is the price metered per interaction with a cap?'

**Free tier** · Yes — A standing free programme gives 1,000 tickets a month with no card, though it runs on a weaker 'free' engine rather than the advanced one.

> Aissist.io offers a Free Program that includes 1,000 free tickets every month. The Free Program runs on the "free" engine, which is sufficient to handle FAQs and basic procedures
> — https://aissist.io/pricing · Pricing FAQ, 'How does the Free Program work?'

**API access** · Partly — The vendor's own documentation audit finds one REST API doc and says it describes Aissist calling your systems, not an interface you drive or pull data from, so programmatic use runs outward only.

> the one REST API doc describes Aissist calling your systems, not you retrieving data from Aissist
> — https://aissist.io/insights/ai-customer-service-vendor-lock-in · Section 'Here is our own finding, and it is not flattering.'

**Bring your own key** · No — Inference is the vendor's, with no bring-your-own-model path recorded in its own lock-in assessment and no key or cloud option published anywhere.

> Moderate — FAQ names OpenAI, Grok, Claude, Gemini; no BYO-model, and the trust center lists only OpenAI
> — https://aissist.io/insights/ai-customer-service-vendor-lock-in · Vendor scoring table, Aissist.io row, 'Model dependency' cell

## Buying it for a team

**A company can buy it** · Yes — Growth is self-serve with no contract and comes with a Standard SLA; the Enterprise tier is sold on contract with a volume discount, a higher SLA and dedicated AI engineer support, and the vendor provides an MSA and a DPA for teams that need them.

> Volume discount, higher SLA, and dedicated AI engineer support, plus customized solutions and premium features.
> — https://aissist.io/pricing · AgentMesh plan cards, Growth and Enterprise columns

**Seat model** · No — Pricing carries no seat component at all: the vendor states its layer is sold per resolution with no platform licence and no seat minimum, in contrast to the seat-based platforms it compares against.

> per resolution, depending on channel as a standalone layer, published on our pricing page, with no platform license or seat minimum.
> — https://aissist.io/industries/aissist-vs-kustomer-ai · Comparison prose on Aissist's own commercial model

**Admin controls** · Yes — Customer-set rules such as refund limits, verification steps and tone are enforced on every conversation and channel, and the vendor frames the guardrails as the customer's to set.

> Refund limits, verification steps and brand tone apply to every conversation, on every channel.
> — https://aissist.io/ · Homepage 'Reliable' section, 'Your rules, enforced' card

**Audit log** · Yes — Audit logs record access and changes across the system and can be produced for a compliance review, which the security page presents as the trail for what the agent did and when.

> Comprehensive audit logs track all access and changes to your system, providing full transparency for compliance requirements.
> — https://aissist.io/security · 'Agentic AI Governance' section, 'Audit Logging' card

## What happens to your code

**Getting out** · No — The home page says Aissist works inside the help desk you already use and moves with you if you switch, so you are never locked in, and the helpdesk stays the record of each conversation; by the vendor's own documentation check it publishes no export path for instructions or sub-agents and no post-termination window.

> Its docs index runs to 50 pages and contains no export page; the one REST API doc describes Aissist calling your systems, not you retrieving data from Aissist
> — https://aissist.io/insights/ai-customer-service-vendor-lock-in · Section 'Here is our own finding, and it is not flattering.'; home page hero for the help desk line

**Certifications** · Partly — ISO 27001 is certified and GDPR compliance comes with a published DPA, but the vendor states plainly that it is SOC 2 aligned rather than SOC 2 certified, with that work still in progress.

> No — Aissist.io is SOC 2 aligned, not SOC 2 certified, and we say so rather than blurring the two.
> — https://aissist.io/security · Security FAQ, 'Are you SOC 2 certified?'

## What changed

- 2026-10-24 — Capabilities · Terms
- 2026-10-22 — Capabilities · Free tier · Price · Terms

## Keeping it current

These facts are re-checked against Aissist.io's own pages and updated when they change. If you represent Aissist.io, sign in and update the listing from your dashboard at https://theagentsindex.com/dashboard. The operations this site accepts from an agent are listed at https://theagentsindex.com/.well-known/agents.json — that catalog needs no key to read, and each entry states its own auth.