CONDITIONAL — official first-party source: Zed's official plans document free use with configured API keys and paid Pro access to hosted models.
Source-BackedZed Agent official pricing page sourceAgenova records Zed Agent as a coding agents product from Zed Industries. Accepted capability evidence: Can configure Agent Panel tools to run automatically or require approval, with rules for trusted, blocked, and sensitive actions. Accepted availability evidence: Zed's official plans document free use with configured API keys and paid Pro access to hosted models.
- Category
- Coding agents
- Purpose taxonomy
- Pending source-backed purpose review
- Pricing
- Freemium
- Deployment
- Desktop
- Source freshness
- Attention required
- Evidence coverage
- 6 of 8 fields
- Execution status
- No published execution evidence
- Last assessed
- 2026-09-09
- Last source scan
- 2026-09-09
- Last material change
- No accepted change recorded
What is Zed Agent?
Zed Agent is the native agent path inside the Zed editor for reading and editing projects, running terminal commands, and using configured tools.
Zed's official documentation describes a native coding Agent that runs in the Agent Panel and Threads Sidebar, can inspect and edit project files, execute terminal commands, use MCP tools, and follow configured profiles and permissions. Zed supports free bring-your-own-key access and paid hosted-model plans. Evaluate model access, tool permissions, terminal approval, data controls and review workflow before adoption.
What is verified—and what remains unknown.
Source-backed now
- PricingZed's official plans document free use with configured API keys and paid Pro access to hosted models.
- CapabilitiesCan configure Agent Panel tools to run automatically or require approval, with rules for trusted, blocked, and sensitive actions.
- DeploymentSupports locally run or self-hosted language models.
- Enterprise controlsSupports allow, deny, and confirmation policies at global and tool-specific levels using matched command, path, or URL patterns.
- Privacy & data policyDiscards code and prompts after each request and does not use them for AI training, except that designated provider models may retain data briefly for safety.
- IntegrationsCan apply per-tool permissions to MCP integrations using server and tool names, including GitHub issue and pull-request tools.
Not yet verified
- Regional availabilityNo accepted public source currently supports this field.
- Execution reliabilityNo publication-approved execution evidence.
Evidence coverage: 75%
6 of 8 decision fields supported. Evidence coverage is not a product score.
- Source-Backed
- 6
- Execution-Verified
- 0
- Intelligence Gap
- 1
- Execution Evidence Gap
- 1
YES — agent tool execution: Can configure Agent Panel tools to run automatically or require approval, with rules for trusted, blocked, and sensitive actions.
Source-BackedZed Agent official integrations sourceYES — model execution: Supports locally run or self-hosted language models.
Source-BackedZed Agent official integrations sourceYES — tool policy controls: Supports allow, deny, and confirmation policies at global and tool-specific levels using matched command, path, or URL patterns.
Source-BackedZed Agent official integrations sourceCONDITIONAL — AI code and prompt data: Discards code and prompts after each request and does not use them for AI training, except that designated provider models may retain data briefly for safety.
Source-BackedZed Agent official privacyDataPolicy sourceYES — MCP tool integrations: Can apply per-tool permissions to MCP integrations using server and tool names, including GitHub issue and pull-request tools.
Source-BackedZed Agent official integrations sourceNo directly supporting official source has passed field-level review.
UnverifiedIntelligence GapNo accepted sourceNo publication-approved execution evidence is available for this Agent.
UnverifiedExecution Evidence GapNo accepted sourceExecution status: No published execution evidence
Documented product facts and Source-Backed claims do not establish execution quality or performance. Only publication-approved execution evidence can create an Execution-Verified state.
Known facts, with boundaries visible.
- Pricing
- FreemiumProduct record
- Deployment
- DesktopProduct record
- Languages
- English
- Provider-listed markets
- GlobalProvider-listed; not verified availability
- Independently verified availability
- Unverified
- Access requirements
- Not recorded
- Payment access
- Not verified
- Network dependency
- Not verified
- Integrations
- YES — MCP tool integrations: Can apply per-tool permissions to MCP integrations using server and tool names, including GitHub issue and pull-request tools.Source-Backed
Capabilities and evidence
Documented product intelligence is not execution evidence or a performance claim. No publication-approved Execution-Verified edge appears unless its governed evidence passes.
No evidence-qualified capability edge is published for this profile. Provider-listed labels are not promoted as verified capabilities.
View full evidence audit9 audited fields · default collapsed
Identity
Source-Backed · 3 matching sources · 2026-09-09Version
Unverified · 0 matching sources · 2026-09-09No accepted supporting source is recorded for this field.
Capabilities
Source-Backed · 1 matching source · 2026-09-06T19:31:18.600ZPricing
Source-Backed · 1 matching source · 2026-09-09T21:54:18.666ZDeployment
Source-Backed · 1 matching source · 2026-09-06T19:28:36.364ZEnterprise controls
Source-Backed · 1 matching source · 2026-09-06T19:31:18.747ZPrivacy & data policy
Source-Backed · 1 matching source · 2026-09-06T19:32:20.419ZIntegrations
Source-Backed · 1 matching source · 2026-09-06T19:31:18.893ZRegional availability
Unverified · 0 matching sources · 2026-09-09No accepted supporting source is recorded for this field.
Official sources
These sources support only the fields named below. A missing source is reported as unknown, not as a negative product claim.