Simulate
Runs thousands of simulated callers with configurable personas (27 voices, 10 languages, 20 background environments) against a voice or chat agent, scores each call on library or custom metrics, and runs from the CLI, API, schedules or GitHub Actions to catch regressions before launch.
What it does
- Evaluation and observability
Sources
- Pricing
-
Starter "100 simulation mins/month", Growth "1,000 simulation mins/month", "Overage per simulation minute" "$0.40 / min" and "$0.25 / min", Enterprise "Custom" "Starting at $4,500/month". The 7-day trial needs a card and is not a free tier.
- Description
-
Official documentation · 21 Oct 2026
"Coval runs thousands of scenarios against your agent, so you know exactly how it performs before a single customer hears it." and "Test against 27 voices, 10 languages, and 20 background environments so nothing surprises you in production."
- What it does
-
Official documentation · 21 Oct 2026
"Choose from a library of resolution, adherence, accuracy, latency, and compliance metrics, or write your own."
- Sold within
-
The pricing page meters this product on its own line ("Simulation minutes / mo" and "Overage per simulation minute") with self-serve signup; the same plans also include "Simulation, monitoring, human review, and observability so your agent can be its best.".
- Name
-
Official documentation · 21 Oct 2026
Menu entry "Simulate"; page label "Product › Simulate".
- URL
-
Official documentation · 21 Oct 2026
Product-menu page "Product › Simulate" (Product menu: Overview, Simulate, Observe, Human Review), in the vendor's llms.txt as "Voice AI Simulation — Test Voice Agents Before Launch".
- Deployment
-
Enterprise plan lists "Private / VPC · Data residency · BYO storage"; the lower plans are hosted.
- Interfaces
-
Official documentation · 21 Oct 2026
"CLI and API first." and "Run evals from your terminal, CI pipeline, or your agent code."; the pricing page (https://www.coval.ai/pricing) lists API, CLI, MCP and skills on every plan.
- Modalities
-
Official documentation · 21 Oct 2026
"Voice-native evaluation." - "Conversation, audio, and tool calls evaluated together in one system, not separately after the fact."
- Intended for
-
Plans are aimed at "For teams starting their agent evaluation journey." (Starter) and "For regulated industries, high-volume, and multi-team orgs." (Enterprise); the Simulate page speaks to engineering teams (https://www.coval.ai/products/simulation).
Also from Coval
-
Observe Platform
Observe scores production voice agent calls, detects performance drift, and provides dashboards and alerts for failures and regressions.
Something wrong here? Send a correction — quote this product id: coval-simulate.