From 9e2ccf408303946b5fec91bd65184808240cf670 Mon Sep 17 00:00:00 2001 From: hongyi-chen Date: Thu, 3 Sep 2026 00:53:37 +0000 Subject: [PATCH 1/3] docs(billing): clarify agent hours per harness and drop the input-pause boundary Follow-up to the agent-hours definition on the platform credits page, incorporating engineering review of the metering behavior: - Describe how much of a run counts per harness. The Warp Agent clock runs while the agent is generating a response and pauses while a command or tool it ran is executing, so a long build or test suite doesn't add to agent hours. Claude Code and Codex bill the whole run, including command execution and time waiting on the model provider. - Drop "pauses for your input" from the stop conditions. It held for the Warp Agent but not for third-party harnesses, which auto-approve in cloud runs. Idle time waiting for a follow-up stays excluded. - Frame agent hours as the meter for platform usage, note that Enterprise contracts set the rate, and say that billing views show credits rather than hours. Verified against warp-server: OpenSegmentForStream/CloseSegmentForStream bound each Warp Agent stream in generate_multi_agent_output.go, while third-party harness runs only close their segment on the terminal transition (task_lifecycle.go, platform.go). Co-Authored-By: Warp --- .../plans-and-billing/platform-credits.mdx | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx b/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx index 8f1e4d18..ea0df124 100644 --- a/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx +++ b/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx @@ -5,7 +5,7 @@ description: >- layer on cloud agent runs and local runs with customer-supplied inference. --- -Platform credits cover Warp's platform layer: run lifecycle, integrations, dashboard, APIs, and observability. They're calculated from agent hours, the time an agent spends working on a task. Platform credits apply to every cloud agent run, including [Warp Factories](/factories/) runs, plus local agent runs on Business and Enterprise plans that use customer-supplied inference. +Platform credits cover Warp's platform layer: run lifecycle, integrations, dashboard, APIs, and observability. They're metered in agent hours, the time an agent spends working on a task. Platform credits apply to every cloud agent run, including [Warp Factories](/factories/) runs, plus local agent runs on Business and Enterprise plans that use customer-supplied inference. Customer-supplied inference is any setup where Warp isn't paying for the model call: [Bring Your Own API Key (BYOK)](/agents/inference/bring-your-own-api-key/), [custom inference endpoints](/agents/inference/custom-inference-endpoint/), [BYOLLM](/enterprise/enterprise-features/bring-your-own-llm/), or Enterprise [team-managed API keys and endpoints](/enterprise/enterprise-features/team-managed-keys-and-endpoints/). @@ -15,9 +15,14 @@ BYOK and custom inference endpoint support are available for individual users an ## How platform credits are measured -Platform credits are calculated from **agent hours**. An agent hour is an hour an agent spends working on a task. The clock starts when the agent starts up in its environment, including any environment setup the run performs, and stops when the run finishes, fails, is cancelled, or pauses for your input. Time spent queued before the agent starts, time waiting on your input, and idle time between turns don't count. +Platform usage is metered in **agent hours**, the time an agent spends working on a task, and billed as platform credits. Time is counted exactly rather than rounded up to whole hours, so a run that works for 15 minutes bills a quarter of an agent hour. See [Warp pricing](https://www.warp.dev/pricing) for current rates; on Enterprise plans, the rate is set by your contract. -Usage is billed on exact time rather than rounded up to whole hours, so a run that works for 15 minutes bills a quarter of an agent hour. See [Warp pricing](https://www.warp.dev/pricing) for current rates. +The clock starts when the agent starts up in its environment, including any environment setup the run performs, and stops when the run finishes, fails, or is cancelled. Time spent queued before the agent starts, and time a run spends idle waiting for your next message, don't count. + +Within that window, how much time counts depends on the [harness](/platform/harnesses/): + +* **Warp Agent** - The clock runs while the agent is generating a response and pauses while a command or tool the agent ran is executing, so a five-minute test suite doesn't add to the run's agent hours. +* **Claude Code and Codex** - The clock runs for the whole run, including command execution and time spent waiting on the model provider. ## The three credit buckets @@ -47,7 +52,7 @@ Whether a run uses platform credits depends on where it runs and who pays for in ## Where platform credits appear -Platform credits appear in the same billing views as the rest of your credit usage. +Platform credits appear in the same billing views as the rest of your credit usage, as credits rather than agent hours. * **Per-user credit totals** - In the Warp app, **Settings** > **Billing and usage** shows platform credits alongside AI credits and compute credits. * **Admin usage breakdown** - For team admins, the Admin Panel billing view shows a per-bucket breakdown for each member and for the team as a whole. Platform credits appear as their own segment in the stacked usage bar. From 43ab397b5d7fb9d20d4d2eaac7a87b3ba8aff5f2 Mon Sep 17 00:00:00 2001 From: hongyi-chen Date: Fri, 4 Sep 2026 22:59:30 +0000 Subject: [PATCH 2/3] docs(billing): rewrite agent-hours section in plain billing terms Replace the "clock starts/stops/runs/pauses" metaphor and casual phrasing ("don't count", "how much time counts") with standard billing vocabulary: a run accrues agent hours, time is or isn't billable, partial hours are billed proportionally. Same facts and boundaries as before; only the register changes. Italicize the defined term on first use, matching the existing convention on the pricing FAQ page. Co-Authored-By: Warp --- .../plans-and-billing/platform-credits.mdx | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx b/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx index ea0df124..a4f450d4 100644 --- a/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx +++ b/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx @@ -15,14 +15,14 @@ BYOK and custom inference endpoint support are available for individual users an ## How platform credits are measured -Platform usage is metered in **agent hours**, the time an agent spends working on a task, and billed as platform credits. Time is counted exactly rather than rounded up to whole hours, so a run that works for 15 minutes bills a quarter of an agent hour. See [Warp pricing](https://www.warp.dev/pricing) for current rates; on Enterprise plans, the rate is set by your contract. +An *agent hour* is one hour an agent spends working on a task. Warp meters platform usage in agent hours and bills that usage as platform credits. Partial hours are billed proportionally, not rounded up: a 15-minute run is a quarter of an agent hour. See [Warp pricing](https://www.warp.dev/pricing) for current rates. Enterprise rates are set by contract. -The clock starts when the agent starts up in its environment, including any environment setup the run performs, and stops when the run finishes, fails, or is cancelled. Time spent queued before the agent starts, and time a run spends idle waiting for your next message, don't count. +A run accrues agent hours from the time the agent starts in its environment, including environment setup, until the run finishes, fails, or is cancelled. Time spent queued before the agent starts isn't billed, and neither is idle time between turns while the run waits for your next message. -Within that window, how much time counts depends on the [harness](/platform/harnesses/): +Within that period, what's billable depends on the [harness](/platform/harnesses/): -* **Warp Agent** - The clock runs while the agent is generating a response and pauses while a command or tool the agent ran is executing, so a five-minute test suite doesn't add to the run's agent hours. -* **Claude Code and Codex** - The clock runs for the whole run, including command execution and time spent waiting on the model provider. +* **Warp Agent** - Time the agent spends generating a response. Execution time for commands and tools the agent runs in its environment, such as a five-minute test suite, isn't billable. +* **Claude Code and Codex** - The entire run, including command execution and time spent waiting on the model provider. ## The three credit buckets @@ -52,7 +52,7 @@ Whether a run uses platform credits depends on where it runs and who pays for in ## Where platform credits appear -Platform credits appear in the same billing views as the rest of your credit usage, as credits rather than agent hours. +Platform credits appear in the same billing views as the rest of your credit usage. Usage is shown in credits, not agent hours. * **Per-user credit totals** - In the Warp app, **Settings** > **Billing and usage** shows platform credits alongside AI credits and compute credits. * **Admin usage breakdown** - For team admins, the Admin Panel billing view shows a per-bucket breakdown for each member and for the team as a whole. Platform credits appear as their own segment in the stacked usage bar. From 0ae38c47cf9ae55e3cc4e8d4c9659e9645295493 Mon Sep 17 00:00:00 2001 From: hongyi-chen Date: Fri, 4 Sep 2026 23:16:47 +0000 Subject: [PATCH 3/3] docs(billing): second pass over the agent-hours pages for consistency and register - Use one phrase for the same fact everywhere: "billed by the agent hour" replaces "calculated from agent hours" (credits, pricing FAQ, platform credits description) and "metered in agent hours" (platform credits intro). - Enterprise billing: split the cloud-agent sentence so the agent-hour definition isn't a mid-sentence parenthetical, note the rate is set by contract, and drop the last "Warp's platform infrastructure" on the page. - Platform credits: the third-party CLI exclusion no longer says "outside Warp", which read as if a CLI run in a Warp pane might be billed. It isn't; only cloud runs on those harnesses are. - Credits: replace "when the conversation is framed around cloud agents versus local agents" with the reason for the alias, and fix a non-parallel list in the usage-prediction sentence. Co-Authored-By: Warp --- .../docs/enterprise/support-and-resources/billing.mdx | 4 ++-- .../support-and-community/plans-and-billing/credits.mdx | 6 +++--- .../plans-and-billing/platform-credits.mdx | 6 +++--- .../plans-and-billing/pricing-faqs.mdx | 2 +- 4 files changed, 9 insertions(+), 9 deletions(-) diff --git a/src/content/docs/enterprise/support-and-resources/billing.mdx b/src/content/docs/enterprise/support-and-resources/billing.mdx index a02f777d..c7af7fb1 100644 --- a/src/content/docs/enterprise/support-and-resources/billing.mdx +++ b/src/content/docs/enterprise/support-and-resources/billing.mdx @@ -31,7 +31,7 @@ Credit allocations vary by contract. Contact your account manager for details on ## Cloud agent billing -Cloud agents consume AI credits for inference, compute credits for the sandbox they run in, and [platform credits](/support-and-community/plans-and-billing/platform-credits/) for the platform layer, billed by the agent hour (the time an agent spends working on a task). Cloud agent runs on self-hosted workers don't consume compute credits but still consume platform credits. See [Credits](/support-and-community/plans-and-billing/credits/) for the full breakdown of how each credit type is metered. +Cloud agents consume AI credits for inference, compute credits for the sandbox they run in, and [platform credits](/support-and-community/plans-and-billing/platform-credits/) for the platform layer. Platform credits are billed by the agent hour, the time an agent spends working on a task, at the rate set by your contract. Cloud agent runs on self-hosted workers don't consume compute credits but still consume platform credits. See [Credits](/support-and-community/plans-and-billing/credits/) for the full breakdown of how each credit type is metered. ### How credits are consumed @@ -43,7 +43,7 @@ For more details, see [Access, Billing, and Identity](/platform/team-access-bill ## BYOLLM billing -When using [Bring Your Own LLM (BYOLLM)](/enterprise/enterprise-features/bring-your-own-llm/), Warp routes requests through your cloud infrastructure ([AWS Bedrock](/enterprise/enterprise-features/byollm-aws-bedrock/) and [Gemini Enterprise Agent Platform (Vertex AI)](/enterprise/enterprise-features/byollm-gemini-enterprise/) today, with Azure Foundry coming soon). Customer-routed BYOLLM inference **consumes no credits** — Warp does not charge credits for requests that execute in your cloud. Inference costs are billed directly to your cloud account. Local agent runs that use BYOLLM still consume [platform credits](/support-and-community/plans-and-billing/platform-credits/) for Warp's platform infrastructure. +When using [Bring Your Own LLM (BYOLLM)](/enterprise/enterprise-features/bring-your-own-llm/), Warp routes requests through your cloud infrastructure ([AWS Bedrock](/enterprise/enterprise-features/byollm-aws-bedrock/) and [Gemini Enterprise Agent Platform (Vertex AI)](/enterprise/enterprise-features/byollm-gemini-enterprise/) today, with Azure Foundry coming soon). Customer-routed BYOLLM inference **consumes no credits** — Warp does not charge credits for requests that execute in your cloud. Inference costs are billed directly to your cloud account. Local agent runs that use BYOLLM still consume [platform credits](/support-and-community/plans-and-billing/platform-credits/). If a BYOLLM request fails and Warp falls back to a direct API model, that fallback request consumes Warp credits at the standard rate. diff --git a/src/content/docs/support-and-community/plans-and-billing/credits.mdx b/src/content/docs/support-and-community/plans-and-billing/credits.mdx index 47af6ebe..4d4cdf2d 100644 --- a/src/content/docs/support-and-community/plans-and-billing/credits.mdx +++ b/src/content/docs/support-and-community/plans-and-billing/credits.mdx @@ -26,7 +26,7 @@ Each interaction consumes **at least one credit**, though more complex interacti For a general breakdown of what factors contribute to how many credits are consumed, see [How are Warp credits calculated?](/support-and-community/plans-and-billing/credits/#how-are-warp-credits-calculated) ::: -Since there's no exact formula for predicting usage, we recommend building an intuitive understanding by experimenting with different prompts, models, and tracking how many credits they consume. +There's no exact formula for predicting usage. Experiment with different prompts and models and track how many credits they consume. ### Tracking your credit usage @@ -121,7 +121,7 @@ These are the most common factors affecting credit usage, though there are other Compute credits cover Warp-hosted compute consumed by an agent run. Cloud agent runs on Warp-hosted workers consume them; local agent runs don't, since they run on your own machine. -Compute credits are sometimes called **cloud agent credits** when the conversation is framed around cloud agents versus local agents. They're the same bucket described from a different angle. +Compute credits are sometimes called **cloud agent credits**, since only cloud agent runs consume them. Both names refer to the same bucket. ### Eligible for compute credits @@ -140,7 +140,7 @@ The following scenarios use compute credits: ## Platform credits -Platform credits are calculated from agent hours, the time an agent spends working on a task. They cover Warp's platform layer (run lifecycle, integrations, dashboard, APIs, and observability) and apply to every cloud agent run, plus local agent runs on Business and Enterprise plans that use customer-supplied inference. +Platform credits are billed by the agent hour, the time an agent spends working on a task. They cover Warp's platform layer (run lifecycle, integrations, dashboard, APIs, and observability) and apply to every cloud agent run, plus local agent runs on Business and Enterprise plans that use customer-supplied inference. ### Eligible for platform credits diff --git a/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx b/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx index a4f450d4..e971d6bb 100644 --- a/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx +++ b/src/content/docs/support-and-community/plans-and-billing/platform-credits.mdx @@ -1,11 +1,11 @@ --- title: Platform credits description: >- - Platform credits are calculated from agent hours and cover Warp's platform + Platform credits are billed by the agent hour and cover Warp's platform layer on cloud agent runs and local runs with customer-supplied inference. --- -Platform credits cover Warp's platform layer: run lifecycle, integrations, dashboard, APIs, and observability. They're metered in agent hours, the time an agent spends working on a task. Platform credits apply to every cloud agent run, including [Warp Factories](/factories/) runs, plus local agent runs on Business and Enterprise plans that use customer-supplied inference. +Platform credits cover Warp's platform layer: run lifecycle, integrations, dashboard, APIs, and observability. They're billed by the agent hour, the time an agent spends working on a task. Platform credits apply to every cloud agent run, including [Warp Factories](/factories/) runs, plus local agent runs on Business and Enterprise plans that use customer-supplied inference. Customer-supplied inference is any setup where Warp isn't paying for the model call: [Bring Your Own API Key (BYOK)](/agents/inference/bring-your-own-api-key/), [custom inference endpoints](/agents/inference/custom-inference-endpoint/), [BYOLLM](/enterprise/enterprise-features/bring-your-own-llm/), or Enterprise [team-managed API keys and endpoints](/enterprise/enterprise-features/team-managed-keys-and-endpoints/). @@ -48,7 +48,7 @@ Whether a run uses platform credits depends on where it runs and who pays for in * **Local agent runs on Free, Build, or Max plans** - Don't use platform credits, whether the run uses Warp-managed or customer-supplied inference. * **Local agent runs on Business or Enterprise with Warp-managed inference** - Don't use platform credits because Warp already charges AI credits for the model call. * **Regular terminal usage** - Shell commands and non-AI Warp features don't consume credits. -* **Third-party agent CLIs run directly** - Running `claude`, `codex`, or another agent CLI outside Warp doesn't consume platform credits. +* **Third-party agent CLIs run directly** - Running `claude`, `codex`, or another agent CLI yourself, whether in a Warp terminal pane or elsewhere, doesn't consume platform credits. ## Where platform credits appear diff --git a/src/content/docs/support-and-community/plans-and-billing/pricing-faqs.mdx b/src/content/docs/support-and-community/plans-and-billing/pricing-faqs.mdx index 5cbb8bc9..98a39ce6 100644 --- a/src/content/docs/support-and-community/plans-and-billing/pricing-faqs.mdx +++ b/src/content/docs/support-and-community/plans-and-billing/pricing-faqs.mdx @@ -162,7 +162,7 @@ Enterprise plans support team-scoped credit pools and per-user spend limits sepa ### How do platform credits factor in? -Warp meters credits across three buckets: **AI credits** (the model call), **compute credits** (the sandbox a cloud agent runs in), and **platform credits** (run lifecycle, integrations, dashboard, APIs, and observability). Platform credits are calculated from agent hours, the time an agent spends working on a task. All three buckets draw from the same pool: your monthly Warp credits first, then [add-on credits](/support-and-community/plans-and-billing/add-on-credits/) once those are exhausted. +Warp meters credits across three buckets: **AI credits** (the model call), **compute credits** (the sandbox a cloud agent runs in), and **platform credits** (run lifecycle, integrations, dashboard, APIs, and observability). Platform credits are billed by the agent hour, the time an agent spends working on a task. All three buckets draw from the same pool: your monthly Warp credits first, then [add-on credits](/support-and-community/plans-and-billing/add-on-credits/) once those are exhausted. Platform credits apply in two situations: