6 reviewed records · 0 benchmark observations · capture dates not recorded
Inspect the evidenceNous Research · TOOL PROFILE
Hermes Agent + Desktop
Persistent, delegated work across projects you own end to end
No editorial score yet — missing data is not a low score
How scoring worksCheck current plans before you buy
Where this comes fromAutonomous coding agents peer board is provisional. No quality rank is published until the declared evidence basket clears. Method AIS-V2
General agent platforms peer board is building cohort. No quality rank is published until the declared evidence basket clears. Method AIS-V2
At a glance
5 facts- Provider
- Nous Research
- Type
- AI tool
- Access
- MIT open source; CLI, TUI, and native desktop app; self-hosted or provider-backed
- Lifecycle
- Active · latest repository release observed Aug 31, 2026
- Pricing notes
- Free to download and use; no account required for the base app. Model, tool, and hosting costs are separate.
Not recorded yet: reviewed evidence records. Gaps stay visible as gaps; nothing is filled in or estimated.
Compared with
Hermes Agent + Desktop sits with Cursor (Autonomous coding agents: Delegate implementation, testing and pull-request work) and Claude Code (Autonomous coding agents: Delegate implementation, testing and pull-request work). Same-job peers are listed together; this profile does not name a winner. Best AI coding tools by job
FAQ
What is Hermes Agent + Desktop?
Hermes Agent + Desktop is an AI tool by Nous Research. Catalogue job: Persistent, delegated work across projects you own end to end. A self-improving AI agent from Nous Research. It is public and MIT licensed, works with multiple model providers, and ships CLI, TUI, and Desktop surfaces with persistent memory, skills, subagents, MCP support, terminal backends, and browser automation.
How does Hermes Agent + Desktop compare with Cursor and Claude Code?
Hermes Agent + Desktop is listed for Autonomous coding agents (“Delegate implementation, testing and pull-request work”) and General agent platforms (“Run persistent work across desktop, terminal, browser, tools and schedules”). Cursor is listed for Autonomous coding agents (“Delegate implementation, testing and pull-request work”); its catalogue best-for is Professional engineers running agent fleets on large codebases. Claude Code is listed for Autonomous coding agents (“Delegate implementation, testing and pull-request work”); its catalogue best-for is Engineers automating multi-file frontend work and PR review. Method AIS-V2 compares peers by job. This page does not pick a winner.
What does Hermes Agent + Desktop cost?
Documented catalogue note: Free to download and use; no account required for the base app. Model, tool, and hosting costs are separate.. Plan terms change. This is a captured note, not a live invoice, and price is not a ranking input.
Follow the source
5 source linksEvidence & conditions
5Source links
6Reviewed records
NoneBenchmark runs
Not recordedLatest capture
- OpenRouter apps marketplace
openrouter.aiDate not recorded - OpenRouter Hermes profile
openrouter.aiDate not recorded - Hermes Desktop
hermes-agent.nousresearch.comDate not recorded - Hermes Agent repository (Nous Research)
github.comDate not recorded - Hermes Agent product site
hermes-agent.nousresearch.comDate not recorded
No comparable benchmark observations are recorded for this entry.
Reviewed records
ev-hermes-capabilities-2026-09-06 capability
- conflict State
- none
- construct
- capability.agent_runtime
- coverage
- 0.82
- entity Id
- hermes-agent
- entity Type
- harness
- entity Version Id
- hermes-agent@2026-09-06
- evidence Type
- capability
- id
- ev-hermes-capabilities-2026-09-06
- independence
- C
- observed At
- 2026-09-06
- output
- capability
- peer Group
- general_agent_platform
- published At
- 2026-08-31
- publisher
- Nous Research
- quality Tier
- moderate
- reproducibility
- B
- review State
- approved
- scope
- Documented Hermes Agent runtime surfaces and workflows
- source Excerpt
- The repository documents CLI, TUI, persistent memory, skills, subagents, scheduling, MCP and seven terminal backends.
- source Type
- first_party_repository
- source Url
- https://github.com/nousresearch/hermes-agent
- state
- verified
- unit
- verified_feature_set
- value
- CLI and TUI agent runtime with persistent memory, skills, subagents, scheduling, MCP and multiple terminal backends
ev-hermes-desktop-2026-09-06 capability
- conflict State
- none
- construct
- capability.desktop_surface
- coverage
- 0.76
- entity Id
- hermes-agent
- entity Type
- product
- entity Version Id
- hermes-agent@2026-09-06
- evidence Type
- capability
- id
- ev-hermes-desktop-2026-09-06
- independence
- C
- observed At
- 2026-09-06
- output
- capability
- peer Group
- general_agent_platform
- published At
- Not documented
- publisher
- Nous Research
- quality Tier
- moderate
- reproducibility
- B
- review State
- approved
- scope
- Desktop availability and supported operating systems
- source Excerpt
- Hermes Desktop for macOS, Windows & Linux; the same agent with a graphical interface.
- source Type
- first_party_product
- source Url
- https://hermes-agent.nousresearch.com/desktop
- state
- verified
- unit
- verified_platform_set
- value
- Native graphical app for macOS 12+, Windows 10/11 and Linux
ev-hermes-openrouter-2026-09-06 adoption
- conflict State
- none
- construct
- adoption.openrouter_usage
- entity Id
- hermes-agent
- entity Type
- product
- entity Version Id
- hermes-agent@2026-09-06
- evidence Type
- adoption
- id
- ev-hermes-openrouter-2026-09-06
- independence
- B
- observed At
- 2026-09-06
- output
- adoption
- peer Group
- general_agent_platform
- published At
- 2026-03-12
- publisher
- OpenRouter
- quality Tier
- moderate
- reproducibility
- B
- review State
- approved
- scope
- OpenRouter Hermes Agent usage profile; adoption only, not quality
- source Excerpt
- Hermes Agent OpenRouter Usage: 50.7T total tokens, daily global rank #1, 455 models used.
- source Type
- independent_platform
- source Url
- https://openrouter.ai/apps/hermes-agent
- state
- source_scoped_rank
- unit
- daily_global_rank
- value
- 1
ev-hermes-price-2026-09-06 price
- conflict State
- none
- construct
- price.software_access
- entity Id
- hermes-agent
- entity Type
- product
- entity Version Id
- hermes-agent@2026-09-06
- evidence Type
- price
- id
- ev-hermes-price-2026-09-06
- independence
- C
- observed At
- 2026-09-06
- peer Group
- general_agent_platform
- published At
- Not documented
- publisher
- Nous Research
- quality Tier
- high
- reproducibility
- A
- review State
- approved
- scope
- Hermes Desktop software access, excluding optional providers and infrastructure
- source Excerpt
- Hermes Desktop is the open-source Hermes Agent under an MIT license, free to download and use. No account is needed to run the base app.
- source Type
- first_party_product
- source Url
- https://hermes-agent.nousresearch.com/desktop
- state
- verified
- unit
- access_model
- value
- Free to download and use; no account required for the base app; model, tool and hosting costs are separate
ev-hermes-quality-gap-2026-09-06 quality
- conflict State
- none
- construct
- quality.comparable_benchmark_basket
- coverage
- 0.1
- entity Id
- hermes-agent
- entity Type
- harness
- entity Version Id
- hermes-agent@2026-09-06
- evidence Type
- quality
- id
- ev-hermes-quality-gap-2026-09-06
- independence
- D
- observed At
- 2026-09-06
- output
- external
- peer Group
- general_agent_platform
- published At
- Not documented
- publisher
- AI Scoreboard review of Nous Research materials
- quality Tier
- low
- reproducibility
- B
- review State
- approved
- scope
- No complete comparable basket found across the six declared General Agent Platforms benchmarks
- source Excerpt
- The reviewed official repository publishes no comparable benchmark score for the declared six-benchmark peer basket.
- source Type
- editorial_gap_review
- source Url
- https://github.com/nousresearch/hermes-agent
- state
- evidence_gap
- unit
- not_available
- value
- Not documented
ev-hermes-user-outcomes-gap-2026-09-06 user_outcomes
- conflict State
- none
- construct
- users.qualified_outcomes
- entity Id
- hermes-agent
- entity Type
- product
- entity Version Id
- hermes-agent@2026-09-06
- evidence Type
- user_outcomes
- id
- ev-hermes-user-outcomes-gap-2026-09-06
- independence
- D
- observed At
- 2026-09-06
- output
- users
- peer Group
- general_agent_platform
- published At
- Not documented
- publisher
- AI Scoreboard review of Nous Research materials
- quality Tier
- low
- reproducibility
- B
- review State
- approved
- scope
- Qualified user outcomes with attributable sample and measurable result
- source Excerpt
- No qualified user-outcome study with an attributable sample and measurable result was found in the reviewed official material.
- source Type
- editorial_gap_review
- source Url
- https://github.com/nousresearch/hermes-agent
- state
- evidence_gap
- unit
- not_available
- value
- Not documented
Source reports, measured results, and editorial assessments are separate evidence types. Source dates do not imply a new benchmark run. Capture dates are shown when recorded; otherwise we say the capture date was not recorded. Recorded dates are compared with this catalogue snapshot (Sep 11, 2026), not today’s date.
Method AIS-V2
How this page is assessed
Quality claims follow Method AIS-V2.
Money never buys rank — commercial relationships cannot change rank, eligibility, evidence quality, or publication timing.
Submit or claim a listing is free and does not skip the editorial queue.
Private · this device only