From 86dedbbe501950132b246818cbf67b3526030598 Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 14:39:55 +0000 Subject: [PATCH 01/13] docs(agents): add Licensing & Usage page and reorder agents manifest --- docs/ai-coder/agents/licensing-usage.md | 49 +++++++++++++++++++++++++ docs/manifest.json | 45 +++++++++++++---------- 2 files changed, 74 insertions(+), 20 deletions(-) create mode 100644 docs/ai-coder/agents/licensing-usage.md diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md new file mode 100644 index 0000000000000..eee3de2eb2118 --- /dev/null +++ b/docs/ai-coder/agents/licensing-usage.md @@ -0,0 +1,49 @@ +# Licensing & Usage + +Coder Agents is licensed differently depending on whether your deployment holds a Community license or a Premium license with a purchased Agent Time allocation. + +## Community licenses + +Community licenses support up to five concurrently active agents. +There is no limit on how long those agents can run or how many tasks they complete over time; they simply queue when more than five agents are active. +This enables individuals and small teams to experiment with Coder Agents at no cost. + +## Premium licenses and Agent Time + +Premium deployments can purchase Agent Time with their Premium license. +Agent Time is shared across the deployment, allowing unlimited agents to run concurrently while consuming from a shared pool of purchased working hours. +This usage-based model is designed for enterprise workloads, where large development teams, background automation, and API-triggered tasks can create highly variable bursts of agent activity without being constrained by a concurrency limit. + +## How Agent Time is measured + +Agent Time is the cumulative duration during which an AI agent is actively processing a task on behalf of the user. +It is measured per interaction step and summed across the duration of a conversation. + +**Includes:** + +- Large language model inference time +- Tool execution (such as file operations, terminal commands, and workspace provisioning) +- Automated error recovery attempts + +**Excludes:** + +- Time the customer spends composing or reviewing messages +- Time between conversation turns when the agent is not processing +- Tool execution delegated to external systems outside the agent's direct control + +Agent Time does not accrue when a conversation is inactive or awaiting user input. +Agent Time is measured with millisecond precision and is rounded down to the nearest minute for billing purposes. + +## Reaching concurrency and usage limits + +### Community concurrency limit + +When a Community license deployment reaches its limit of five concurrently active agents, any additional agents are placed in a queue. +Queued agents are picked up automatically as soon as capacity frees up, so no work is lost and no action is required from the user. + +### Premium Agent Time exhaustion + +When a Premium deployment exhausts its purchased Agent Time, the deployment may be provisioned to fall back to the Community concurrency model until more Agent Time is purchased. +In this state, agents are limited to five concurrent active agents, and any additional agents are queued until capacity is available. + +Deployment administrators see an in-app soft warning message as the deployment approaches its maximum allotted Agent Time, so they can purchase additional Agent Time before the concurrency fallback takes effect. diff --git a/docs/manifest.json b/docs/manifest.json index c4aa0ee5549cb..4980584c09b20 100644 --- a/docs/manifest.json +++ b/docs/manifest.json @@ -1033,31 +1033,11 @@ "description": "Enable Coder Agents, prepare your deployment, and run your first Coder Agent", "path": "./ai-coder/agents/getting-started.md" }, - { - "title": "Search Syntax", - "description": "Filter Coder Agents conversations by title, status, and linked pull requests with search syntax.", - "path": "./ai-coder/agents/chat-search-syntax.md" - }, { "title": "Architecture", "description": "Learn how the Coder Agents control-plane agent communicates with your workspaces.", "path": "./ai-coder/agents/architecture.md" }, - { - "title": "Chat Sharing", - "description": "Give other users and groups read-only access to a Coder Agents conversation with chat sharing.", - "path": "./ai-coder/agents/chat-sharing.md" - }, - { - "title": "Models", - "description": "Configure the LLM providers and models that power Coder Agents from Admin settings.", - "path": "./ai-coder/agents/models.md" - }, - { - "title": "Tools", - "description": "How Coder Agents uses tool calls to select templates and create workspaces", - "path": "./ai-coder/agents/tools/index.md" - }, { "title": "Platform Controls", "description": "Learn how platform teams control Coder Agents behavior, models, spend, and policies.", @@ -1124,6 +1104,31 @@ "description": "Extend Coder Agents with custom skills and MCP tools defined in your workspace templates.", "path": "./ai-coder/agents/extending-agents.md" }, + { + "title": "Models", + "description": "Configure the LLM providers and models that power Coder Agents from Admin settings.", + "path": "./ai-coder/agents/models.md" + }, + { + "title": "Tools", + "description": "How Coder Agents uses tool calls to select templates and create workspaces", + "path": "./ai-coder/agents/tools/index.md" + }, + { + "title": "Chat Sharing", + "description": "Give other users and groups read-only access to a Coder Agents conversation with chat sharing.", + "path": "./ai-coder/agents/chat-sharing.md" + }, + { + "title": "Search Syntax", + "description": "Filter Coder Agents conversations by title, status, and linked pull requests with search syntax.", + "path": "./ai-coder/agents/chat-search-syntax.md" + }, + { + "title": "Licensing & Usage", + "description": "Licensing and usage details for Coder Agents.", + "path": "./ai-coder/agents/licensing-usage.md" + }, { "title": "Tasks to Chats API Migration", "description": "Migrate your integrations from the Coder Tasks API to the newer Coder Agents Chats API.", From 5822237148953cdcca14d49a712a2b9f0bca8d7a Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 14:46:53 +0000 Subject: [PATCH 02/13] docs(agents): rename Premium to AI Premium and refine licensing copy --- docs/ai-coder/agents/licensing-usage.md | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index eee3de2eb2118..02b4c9f6c7bf9 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -1,16 +1,16 @@ # Licensing & Usage -Coder Agents is licensed differently depending on whether your deployment holds a Community license or a Premium license with a purchased Agent Time allocation. +Coder Agents is licensed differently depending on whether your deployment holds a Community license or an AI Premium license with a purchased Agent Time allocation. ## Community licenses -Community licenses support up to five concurrently active agents. +Community licenses support up to five concurrently active agents deployment-wide. There is no limit on how long those agents can run or how many tasks they complete over time; they simply queue when more than five agents are active. This enables individuals and small teams to experiment with Coder Agents at no cost. -## Premium licenses and Agent Time +## AI Premium licenses and Agent Time -Premium deployments can purchase Agent Time with their Premium license. +AI Premium licenses include a customizable amount of Agent Time. Agent Time is shared across the deployment, allowing unlimited agents to run concurrently while consuming from a shared pool of purchased working hours. This usage-based model is designed for enterprise workloads, where large development teams, background automation, and API-triggered tasks can create highly variable bursts of agent activity without being constrained by a concurrency limit. @@ -41,9 +41,9 @@ Agent Time is measured with millisecond precision and is rounded down to the nea When a Community license deployment reaches its limit of five concurrently active agents, any additional agents are placed in a queue. Queued agents are picked up automatically as soon as capacity frees up, so no work is lost and no action is required from the user. -### Premium Agent Time exhaustion +### AI Premium Agent Time exhaustion -When a Premium deployment exhausts its purchased Agent Time, the deployment may be provisioned to fall back to the Community concurrency model until more Agent Time is purchased. +When an AI Premium deployment exhausts its purchased Agent Time, the deployment may be provisioned to fall back to the Community concurrency model until more Agent Time is purchased. In this state, agents are limited to five concurrent active agents, and any additional agents are queued until capacity is available. Deployment administrators see an in-app soft warning message as the deployment approaches its maximum allotted Agent Time, so they can purchase additional Agent Time before the concurrency fallback takes effect. From 55500265b2a760eb47043bc3f1bb08f817a67e6f Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 15:01:46 +0000 Subject: [PATCH 03/13] docs: escape ampersand in manifest title for gen check --- docs/manifest.json | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/manifest.json b/docs/manifest.json index 4980584c09b20..9e5c1977ce7db 100644 --- a/docs/manifest.json +++ b/docs/manifest.json @@ -1125,7 +1125,7 @@ "path": "./ai-coder/agents/chat-search-syntax.md" }, { - "title": "Licensing & Usage", + "title": "Licensing \u0026 Usage", "description": "Licensing and usage details for Coder Agents.", "path": "./ai-coder/agents/licensing-usage.md" }, From e5a8359bb4715e7c26c2a958115d30ce6bf93d63 Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 13:53:27 -0400 Subject: [PATCH 04/13] Update docs/ai-coder/agents/licensing-usage.md Co-authored-by: Nick Vigilante --- docs/ai-coder/agents/licensing-usage.md | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index 02b4c9f6c7bf9..a5cc671ca57cd 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -1,4 +1,6 @@ -# Licensing & Usage +--- +title: Licensing & Usage +--- Coder Agents is licensed differently depending on whether your deployment holds a Community license or an AI Premium license with a purchased Agent Time allocation. From 048501da3d44edb750766eec2093e04c9cb4f840 Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 13:53:55 -0400 Subject: [PATCH 05/13] Update docs/ai-coder/agents/licensing-usage.md Co-authored-by: Nick Vigilante --- docs/ai-coder/agents/licensing-usage.md | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index a5cc671ca57cd..f5c9d2c609924 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -7,7 +7,8 @@ Coder Agents is licensed differently depending on whether your deployment holds ## Community licenses Community licenses support up to five concurrently active agents deployment-wide. -There is no limit on how long those agents can run or how many tasks they complete over time; they simply queue when more than five agents are active. +Coder doesn't limit how long those agents can run or how many tasks they complete over time. +Agents will queue when more than five agents are active at a time. This enables individuals and small teams to experiment with Coder Agents at no cost. ## AI Premium licenses and Agent Time From 38959f8af68098f2b6515728d7353703818d9b49 Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 13:54:25 -0400 Subject: [PATCH 06/13] Update docs/ai-coder/agents/licensing-usage.md Co-authored-by: Nick Vigilante --- docs/ai-coder/agents/licensing-usage.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index f5c9d2c609924..1a97acbf695d4 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -9,7 +9,7 @@ Coder Agents is licensed differently depending on whether your deployment holds Community licenses support up to five concurrently active agents deployment-wide. Coder doesn't limit how long those agents can run or how many tasks they complete over time. Agents will queue when more than five agents are active at a time. -This enables individuals and small teams to experiment with Coder Agents at no cost. +With this agent pool, individuals and small teams can experiment with Coder Agents at no cost. ## AI Premium licenses and Agent Time From e39ea28fefea31133a1841a1651157f1f176ca53 Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 13:54:34 -0400 Subject: [PATCH 07/13] Update docs/ai-coder/agents/licensing-usage.md Co-authored-by: Nick Vigilante --- docs/ai-coder/agents/licensing-usage.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index 1a97acbf695d4..95e4cf93b4638 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -37,7 +37,7 @@ It is measured per interaction step and summed across the duration of a conversa Agent Time does not accrue when a conversation is inactive or awaiting user input. Agent Time is measured with millisecond precision and is rounded down to the nearest minute for billing purposes. -## Reaching concurrency and usage limits +## Concurrency and usage limits ### Community concurrency limit From b13a214728987250350f9fd63082330dda214cde Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 13:54:42 -0400 Subject: [PATCH 08/13] Update docs/ai-coder/agents/licensing-usage.md Co-authored-by: Nick Vigilante --- docs/ai-coder/agents/licensing-usage.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index 95e4cf93b4638..45dca46af1d29 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -19,7 +19,7 @@ This usage-based model is designed for enterprise workloads, where large develop ## How Agent Time is measured -Agent Time is the cumulative duration during which an AI agent is actively processing a task on behalf of the user. +Agent Time is the cumulative duration during which an AI agent is actively processing a task for the user. It is measured per interaction step and summed across the duration of a conversation. **Includes:** From 10db66424ce52a86ae63feebc574a0f8b04ae4e0 Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 13:55:06 -0400 Subject: [PATCH 09/13] Update docs/ai-coder/agents/licensing-usage.md Co-authored-by: Nick Vigilante --- docs/ai-coder/agents/licensing-usage.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index 45dca46af1d29..087d475384cdf 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -39,6 +39,8 @@ Agent Time is measured with millisecond precision and is rounded down to the nea ## Concurrency and usage limits +Coder handles concurrency and usage limits differently, depending on the license you use for Coder. + ### Community concurrency limit When a Community license deployment reaches its limit of five concurrently active agents, any additional agents are placed in a queue. From 493a617624eda373dd5f046e0eb25c117e34c81c Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 13:55:25 -0400 Subject: [PATCH 10/13] Update docs/ai-coder/agents/licensing-usage.md Co-authored-by: Nick Vigilante --- docs/ai-coder/agents/licensing-usage.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index 087d475384cdf..c539b962e3832 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -44,7 +44,7 @@ Coder handles concurrency and usage limits differently, depending on the license ### Community concurrency limit When a Community license deployment reaches its limit of five concurrently active agents, any additional agents are placed in a queue. -Queued agents are picked up automatically as soon as capacity frees up, so no work is lost and no action is required from the user. +As soon as an active agent completes its task, the next queued agent begins its work, so no work is lost and no action is required from the user. ### AI Premium Agent Time exhaustion From f50a803602bbeeda4a3759d0c98808480d331340 Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 13:55:43 -0400 Subject: [PATCH 11/13] Update docs/ai-coder/agents/licensing-usage.md Co-authored-by: Nick Vigilante --- docs/ai-coder/agents/licensing-usage.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index c539b962e3832..550c24674356b 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -49,6 +49,6 @@ As soon as an active agent completes its task, the next queued agent begins its ### AI Premium Agent Time exhaustion When an AI Premium deployment exhausts its purchased Agent Time, the deployment may be provisioned to fall back to the Community concurrency model until more Agent Time is purchased. -In this state, agents are limited to five concurrent active agents, and any additional agents are queued until capacity is available. +In this state, your Coder deployment is limited to five concurrent active agents, and any additional agents are queued until capacity is available. Deployment administrators see an in-app soft warning message as the deployment approaches its maximum allotted Agent Time, so they can purchase additional Agent Time before the concurrency fallback takes effect. From c6c13ad90acf202088e097bdf4c12963134d6590 Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 13:56:12 -0400 Subject: [PATCH 12/13] Update docs/ai-coder/agents/licensing-usage.md Co-authored-by: Nick Vigilante --- docs/ai-coder/agents/licensing-usage.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index 550c24674356b..9a90c2169c75a 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -51,4 +51,4 @@ As soon as an active agent completes its task, the next queued agent begins its When an AI Premium deployment exhausts its purchased Agent Time, the deployment may be provisioned to fall back to the Community concurrency model until more Agent Time is purchased. In this state, your Coder deployment is limited to five concurrent active agents, and any additional agents are queued until capacity is available. -Deployment administrators see an in-app soft warning message as the deployment approaches its maximum allotted Agent Time, so they can purchase additional Agent Time before the concurrency fallback takes effect. +Coder sends deployment administrators an in-app soft warning message as the deployment approaches its maximum allotted Agent Time, so they can purchase additional Agent Time before the concurrency fallback takes effect. From ded112000ce8938aaa3ff1886684cd665670ff8d Mon Sep 17 00:00:00 2001 From: Matt Vollmer Date: Tue, 18 Aug 2026 18:51:37 +0000 Subject: [PATCH 13/13] docs: remove fallback detail from AI Premium Agent Time exhaustion --- docs/ai-coder/agents/licensing-usage.md | 3 --- 1 file changed, 3 deletions(-) diff --git a/docs/ai-coder/agents/licensing-usage.md b/docs/ai-coder/agents/licensing-usage.md index 9a90c2169c75a..047d4aaf137b8 100644 --- a/docs/ai-coder/agents/licensing-usage.md +++ b/docs/ai-coder/agents/licensing-usage.md @@ -48,7 +48,4 @@ As soon as an active agent completes its task, the next queued agent begins its ### AI Premium Agent Time exhaustion -When an AI Premium deployment exhausts its purchased Agent Time, the deployment may be provisioned to fall back to the Community concurrency model until more Agent Time is purchased. -In this state, your Coder deployment is limited to five concurrent active agents, and any additional agents are queued until capacity is available. - Coder sends deployment administrators an in-app soft warning message as the deployment approaches its maximum allotted Agent Time, so they can purchase additional Agent Time before the concurrency fallback takes effect.