What did Hlido score Avots MCP?

Avots MCP scored 72/100 (STEADY) in Hlido's independent, hands-on review.

Does any vendor pay Hlido for placement?

No. Hlido takes no money from the agents it rates — scoring weights stay private and the evidence behind every verdict is public.

Avots MCP

Name: Avots MCP review
Item: Avots MCP
Rating: 72
Author: Hlido Editor

Infrastructure · tested 2026-08-05 · re-test due 2026-11-05 · by the Hlido desk, not the vendor

In short: A shipped multi-model gateway reachable over MCP — concrete, specific, and documented for six clients, with the usual gateway caveats about depending on a middleman.

5 PASS · 0 FAIL of 5 public-surface claims

Visit Avots MCP → scorecard.json

Quick answer

Avots MCP scores 72/100 (STEADY) on Hlido’s independent, hands-on test (reviewed 2026-08-05). STEADY (72) because this is shipped and specific rather than promised: sixteen named tools with free/paid status, a precisely stated transport (streamable HTTP, MCP spec 2025-06-18, bearer auth), OAuth for Claude web, an

Hlido Editor · 2026-08-05

Avots exposes 40+ AI models through a single MCP endpoint, so a Claude, Cursor, Cline or openclaw session can call GPT-5, DeepSeek, Sonar, Veo, Kling, ElevenLabs and others without leaving the client. What makes this reviewable rather than aspirational is specificity: sixteen named tools are listed with what each does and which are free (check_balance, list_models, check_job, calendar and avatar listing), the transport is stated precisely (streamable HTTP, MCP spec 2025-06-18, bearer auth), OAuth is supported for Claude web so there is no token copy-paste, and there is a per-client setup guide for six different clients. The generative range is unusually broad for one endpoint — chat, image, video, audio, face swap, talking avatars, lipsync, vlogs, montage. The trade-offs are inherent to the pattern: every call is routed through a third party, so the vendor sees prompts and outputs and becomes a single point of failure; the surface shows no pricing detail, uptime record, rate limits, or data-handling policy in what we captured; and there is no adoption evidence or third-party validation. A two-step confirmation on video generation is a sensible guard against expensive accidental calls and worth noting.

Why STEADY

STEADY (72) because this is shipped and specific rather than promised: sixteen named tools with free/paid status, a precisely stated transport (streamable HTTP, MCP spec 2025-06-18, bearer auth), OAuth for Claude web, and setup guides for six clients. Held below VITAL because the gateway pattern concentrates prompts, outputs and availability in one third party, and the captured surface shows no pricing detail, uptime record, rate limits or data-handling policy — the disclosures that would make that concentration acceptable.

Public-surface checklist

PASS Homepage loads (required)
PASS Primary value prop (required) — 'Use avots inside your favourite AI client'
PASS Cta present (required) — 'Get my MCP key'
PASS Pricing or access — A pricing page exists in navigation; no figures on the captured MCP page
PASS Evidence or demo — Concrete per-client setup guides and a full tool list; no third-party proof

What we saw

2 screenshots captured by the Hlido engine during the reviewed run (run-e39c90cd0b4c4745-avots-ai). Our own captures — not vendor marketing material.

Avots MCP — run screenshot 1 (home.png) — home.png

Avots MCP — run screenshot 2 (page_faq.png) — page_faq.png

What it does well

Sixteen tools listed by name with function and free/paid status — no ambiguity about what you get
One MCP endpoint reaches 40+ models across chat, image, video, audio and avatar generation
Precise transport disclosure: streamable HTTP, MCP spec 2025-06-18, Authorization bearer
OAuth support for Claude.ai web — connect without copy-pasting a token
Per-client setup guides for Claude web, Claude Desktop, Claude Code, Cursor, Cline and openclaw
Two-step confirmation on video generation — a sensible guard against costly accidental calls

What it fails at

Every request routes through a third party that sees prompts and outputs; no data-handling or retention policy on the captured surface
Single point of failure — if the gateway is down, all 40+ models are unreachable from the client
No pricing detail, rate limits, or uptime/reliability record published on the MCP page
No adoption numbers, named users, or third-party validation
Model availability depends on upstream providers; nothing states what happens when one is withdrawn

Best for

Developers who want many models available inside one MCP client without managing separate provider keys
Multimodal work (image, video, audio, avatar) reachable from a chat or IDE session
Quick experimentation across models where per-provider onboarding is the friction

Not recommended for

Work with confidential prompts where routing through an intermediary is unacceptable
Production systems needing an SLA, published rate limits, or direct provider relationships
Buyers who require a stated data-retention and processing policy before sending anything

Pricing & access

Pricing findable on the public surfacePASS A pricing page exists in navigation; no figures on the captured MCP page (tested 2026-08-05)

Derived from Hlido-held evidence only (engine checklist + editorial text); quotes are verbatim from the scorecard; not vendor-supplied; re-derived daily. Verify current prices on the vendor's pricing page. Last verified 2026-08-05.

Related agents

Agent relevance

API MCP Behavioral-testable

Agentic-Commerce Readiness 74/100 · INTEGRABLE

Independent readiness for agent delegation & transaction. How it’s scored · check live

MCP-first: a hosted streamable-HTTP endpoint with bearer auth and OAuth, documented for six MCP clients. The product exists to be called by an agent inside another client.

Agent-friendly score: 9/10

Evidence

Single MCP endpoint at https://mcp.avots.ai/ exposing 40+ models — source (2026-08-05) verified
16 named tools with per-tool descriptions and free/paid marking — source (2026-08-05) verified
Streamable HTTP transport, MCP spec 2025-06-18, bearer authorization — source (2026-08-05) verified
OAuth connection flow for Claude.ai web (no token copy-paste) — source (2026-08-05) verified
Pricing detail, rate limits, uptime record, data-handling policy — source (2026-08-05)

scorecard.json · transparency passport · registry · methodology

More: compare agents · best of · developer tools · incident registry

Verdict by Hlido Editor, our automated editorial system · Method: public-surface-tier-1+editorial-narrative-v2 · Methodology version 2026.05 · Next review due 2026-11-05

How this page was produced. The scores, claim verdicts and evidence come from automated hands-on testing of the product’s public surface. The written analysis is drafted by an AI system, and pages publish without a person reviewing each one. Hlido publishes this record and answers for it — tell us if anything here is wrong and we will correct it.

Embed this trust badge

Live, always-current independent score — free to embed on your site or README. No vendor pays for placement.

Markdown

[![Hlido trust score](https://hlido.eu/badge/avotsai-avots-mcp.svg)](https://hlido.eu/check/?agent=avotsai-avots-mcp)

HTML

<a href="https://hlido.eu/check/?agent=avotsai-avots-mcp"><img src="https://hlido.eu/badge/avotsai-avots-mcp.svg" alt="Hlido trust score"></a>