jprdonnelly/docs
1
0
Fork 0
mirror of synced 2025-12-19 09:57:42 -05:00
Code Issues Projects Releases Wiki Activity

GitHub Models Billing for GitHub Paper Customers [Public Preview] (#56168)

Browse Source 
Co-authored-by: Felicity Chapman <felicitymay@github.com>
Co-authored-by: Sunbrye Ly <56200261+sunbrye@users.noreply.github.com>
Co-authored-by: sunbrye <sunbrye@github.com>
Co-authored-by: mc <42146119+mchammer01@users.noreply.github.com>
This commit is contained in:
Vanessa
2025-06-26 03:23:11 +10:00
committed by GitHub
parent 48f820b19b
commit c1c31fa621
16 changed files with 227 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
content/admin/data-residency/feature-overview-for-github-enterprise-cloud-with-data-residency.md
View File

@@ -30,6 +30,7 @@ The following features are currently unavailable on {% data variables.enterprise
| {% data variables.product.prodname_copilot_short %} Metrics API | Currently unavailable | [AUTOTITLE](/rest/copilot/copilot-metrics) |
| {% data variables.copilot.copilot_coding_agent %} | Currently unavailable | [AUTOTITLE](/copilot/using-github-copilot/coding-agent/about-assigning-tasks-to-copilot) |
| Restricting {% data variables.product.prodname_actions %} policies to verified creators | Currently unavailable | [AUTOTITLE](/admin/enforcing-policies/enforcing-policies-for-your-enterprise/enforcing-policies-for-github-actions-in-your-enterprise#allow-enterprise-and-select-non-enterprise-actions-and-reusable-workflows) |
| {% data variables.product.prodname_github_models %} | Currently unavailable | [AUTOTITLE](/github-models/about-github-models) |
| Some features currently in {% data variables.release-phases.public_preview %} or {% data variables.release-phases.private_preview %} | Certain features that are in a preview phase on {% data variables.product.prodname_dotcom_the_website %} may not be available on {% data variables.enterprise.data_residency_site %} | |
## Features that work differently

142
content/billing/managing-billing-for-your-products/about-billing-for-github-models.md Normal file
View File

@@ -0,0 +1,142 @@
---
title: About billing for GitHub Models
shortTitle: GitHub Models
intro: If you want to use {% data variables.product.prodname_github_models %} beyond the free usage included in your account, you can choose to opt in to paid usage.
versions:
feature: github-models
topics:
- Enterprise
- Billing
allowTitleToDifferFromFilename: true
---
> [!NOTE]
> * {% data variables.product.prodname_github_models %} for organizations and repositories is in {% data variables.release-phases.public_preview %} and subject to change.
> * Billing for {% data variables.product.prodname_github_models %} is separate from billing for {% data variables.product.prodname_copilot %}. For more information about how models in {% data variables.product.prodname_copilot %} are billed, see [AUTOTITLE](/billing/managing-billing-for-your-products/about-billing-for-github-copilot).
## About {% data variables.product.prodname_github_models %} billing on {% data variables.product.github %}
Billing for {% data variables.product.prodname_github_models %} is designed to be flexible and to allow you to use your preferred model providers, while also providing the ability to control your spending. Model usage is powered by the Azure OpenAI Service and billed through {% data variables.product.github %} using the same global pay-as-you-go pricing as [Azure OpenAI Service](https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/).
Alternatively, you can bring your own API keys (BYOK) from different providers, such as OpenAI or Azure AI, see [AUTOTITLE](/github-models/github-models-at-scale/set-up-custom-model-integration-models-byok). Model inference runs directly through your provider, and usage is billed and tracked through your provider account. For example, Azure API key usage is billed and tracked through your Azure Subscription ID.
Each {% data variables.product.github %} account receives a certain amount of included free but rate-limited usage of {% data variables.product.prodname_github_models %}, see [Rate limits](/github-models/use-github-models/prototyping-with-ai-models#rate-limits).
The use of {% data variables.product.prodname_github_models %} from the {% data variables.product.prodname_marketplace %} catalog is subject to the free rate limits, and is not billed.
## Available models
The table below lists the currently available models. Each model supported by {% data variables.product.prodname_github_models %} has an input and output multiplier that determines the number of token units for each request. For more information on token units, see [Pricing for paid usage](#pricing-for-paid-usage).
| Model name | Input multiplier | Cached input multiplier | Output multiplier | Input price (per 1M token units) | Cached input price (per 1M token units) | Output price (per 1M token units) |
| --------------------------------- | ---------------- | ----------------------- | ----------------- |----------------------------------|-----------------------------------------|-----------------------------------|
| OpenAI GPT-4o | 0.25 | 0.125 | 1.0 | $2.50 | $1.25 | $10.00 |
| OpenAI GPT-4o mini | 0.015 | 0.0075 | 0.06 | $0.15 | $0.08 | $0.60 |
| OpenAI GPT-4.1-mini | 0.04 | 0.01 | 0.16 | $0.40 | $0.10 | $1.60 |
| OpenAI GPT-4.1 | 0.2 | 0.05 | 0.8 | $2.00 | $0.50 | $8.00 |
| Phi-3.5-MoE instruct (128k) | 0.016 | N/A | 0.064 | $0.16 | N/A | $0.64 |
| Phi-3.5-mini instruct (128k) | 0.013 | N/A | 0.052 | $0.13 | N/A | $0.52 |
| Phi-3.5-vision instruct (128k) | 0.013 | N/A | 0.052 | $0.13 | N/A | $0.52 |
| Phi-3-medium instruct (4k) | 0.017 | N/A | 0.068 | $0.17 | N/A | $0.68 |
| Phi-3-medium instruct (128k) | 0.017 | N/A | 0.068 | $0.17 | N/A | $0.68 |
| Phi-3-mini instruct (4k) | 0.013 | N/A | 0.052 | $0.13 | N/A | $0.52 |
| Phi-3-mini instruct (128k) | 0.013 | N/A | 0.052 | $0.13 | N/A | $0.52 |
| Phi-3-small instruct (8k) | 0.015 | N/A | 0.06 | $0.15 | N/A | $0.60 |
| Phi-3-small instruct (128k) | 0.015 | N/A | 0.06 | $0.15 | N/A | $0.60 |
| Phi-4 | 0.0125 | N/A | 0.05 | $0.13 | N/A | $0.50 |
| Phi-4-mini-instruct | 0.0075 | N/A | 0.03 | $0.08 | N/A | $0.30 |
| Phi-4-multimodal-instruct | 0.008 | N/A | 0.032 | $0.08 | N/A | $0.32 |
| DeepSeek-R1 | 0.135 | N/A | 0.54 | $1.35 | N/A | $5.40 |
| DeepSeek-R1-0528 | 0.135 | N/A | 0.54 | $1.35 | N/A | $5.40 |
| DeepSeek-V3-0324 | 0.114 | N/A | 0.456 | $1.14 | N/A | $4.56 |
| MAI-DS-R1 | 0.135 | N/A | 0.54 | $1.35 | N/A | $5.40 |
| Grok 3 Mini | 0.025 | N/A | 0.127 | $0.25 | N/A | $1.27 |
| Grok 3 | 0.3 | N/A | 1.5 | $3.00 | N/A | $15.00 |
| Llama 4 Maverick 17B Instruct FP8 | 0.025 | N/A | 0.1 | $0.25 | N/A | $1.00 |
| Llama-3.3-70B-Instruct | 0.071 | N/A | 0.071 | $0.71 | N/A | $0.71 |
For accounts that use a custom model with a third-party model provider, billing is managed through the provider and is subject to the provider’s own pricing. For more information, see the official pricing documentation for your chosen provider.
## Opting in to paid usage
> [!NOTE] Once you opt in to paid usage, you will have access to production grade rate limits and be billed for all usage thereafter. For more information about these rate limits, see [Azure AI Foundry Models quotas and limits](https://learn.microsoft.com/en-us/azure/ai-foundry/model-inference/quotas-limits) in the Azure documentation.
Enterprises and organizations can opt in to paid usage to access expanded model capabilities, including increased request allowances and larger context windows. You can manage their spending by setting a budget.
By default, organizations and personal accounts that opt in to paid usage have a spending limit of $0 US dollars (USD) until the budget is increased.
For more information, see [AUTOTITLE](/billing/managing-your-billing/using-budgets-control-spending).
### How do I pay for {% data variables.product.prodname_github_models %}?
{% data variables.product.prodname_github_models %} usage can be paid for by one or more of the following methods:
* For enterprises, organizations, or personal accounts directly billed by {% data variables.product.github %}, the billing of {% data variables.product.prodname_github_models %} is based on your metered usage for each billing period, and pricing varies by the number of model requests, tokens, and the model multiplier.
* For invoiced accounts, contact {% data variables.contact.contact_enterprise_sales %} to discuss billing for {% data variables.product.prodname_github_models %} usage.
* When using custom models with your own API keys, the billing of {% data variables.product.prodname_github_models %} is based on the model provider's pricing, and not {% data variables.product.github %}.
* Accounts with an existing Azure subscription can use that subscription to pay for model inference by bringing their own API key for custom models. In this case, billing is based on the model provider’s pricing and is managed through the Azure subscription. See [AUTOTITLE](/github-models/github-models-at-scale/set-up-custom-model-integration-models-byok).
> [!TIP]
> You are considered to be directly billed by {% data variables.product.github %} if you pay for {% data variables.product.github %} using a credit card, PayPal, or by invoice.
## Pricing for paid usage
{% data variables.product.prodname_github_models %} pricing is based on the number of token units used, at a fixed price of $0.00001 USD per token unit.
A token unit is calculated by multiplying the number of input and output tokens by their respective model multipliers. All model usage, regardless of the underlying provider or model, is measured in token units. While some providers display prices per 1,000 or per 1,000,000 tokens, {% data variables.product.prodname_github_models %} standardizes billing to the token unit level. This means you are billed using a single SKU and a unified price per token unit, no matter which supported model you use.
Your cost is calculated by multiplying the number of token units you use by the unified token unit price.
At the end of your billing cycle, {% data variables.product.github %} calculates the cost of token units used, starting from your first request after opting in to paid usage.
The number of model requests and tokens you have used are reset every billing cycle.
### Sample models cost calculation
The following table displays how the total cost is calculated for a request using OpenAI GPT-4o:
| Model | Input tokens used | Output tokens used | Input multiplier | Output multiplier | Total token units | Price per token unit | Total cost |
|---------------| ----------------- | ------------------ | ---------------- | ----------------- |-------------------|----------------------|------------|
| OpenAI GPT-4o | 1,000,000 | 1,000,000 | 0.25 | 1 | 1,250,000 | $0.00001 | $12.50 |
### Calculating model costs
The following steps demonstrate how the total cost is calculated:
1. **Calculate input tokens:**
Multiply the number of input tokens by the input multiplier.
`1,000,000 tokens × 0.25 = 250,000 input token units`
1. **Calculate billable output tokens:**
Multiply the number of output tokens by the output multiplier.
`1,000,000 tokens × 1 = 1,000,000 output token units`
1. **Add billable tokens:**
Add the billable input and output tokens.
`250,000 (input) + 1,000,000 (output) = 1,250,000 total token units`
1. **Charges by type:**
* **Input charge:** `250,000 × $0.00001 = $2.50`
* **Output charge:** `1,000,000 × $0.00001 = $10.00`
1. **Calculate the total cost:**
Multiply the total token units by the token unit price.
`1,250,000 × $0.00001 = $12.50 for this request`
## Opting out of paid usage
{% data variables.product.prodname_github_models %} billing is disabled by default for enterprises and organizations. An enterprise must enable paid usage before any organization within it can opt in to billing. Once an enterprise or organization has opted in to paid usage, the billing is enabled for all repositories owned by the enterprise or organization, including repositories owned by {% data variables.product.prodname_emus %} (EMUs).
Enterprises control whether paid usage for {% data variables.product.prodname_github_models %} is available to organizations they own. An enterprise must opt in to paid usage before any organization within it can opt in and be billed. If the enterprise disables billing, no organizations in the enterprise can enable or use paid features.
Organizations can opt out of paid usage by setting a budget, or instead choose to disable {% data variables.product.prodname_github_models %} entirely by disabling the product in the organization's settings, see [Controlling model usage in your organization](/github-models/github-models-at-scale/manage-models-at-scale#controlling-model-usage-in-your-organization).
If an enterprise has opted in to billing for {% data variables.product.prodname_github_models %}, but an organization within the enterprise has opted out of billing, then paid {% data variables.product.prodname_github_models %} usage is disabled for the organization, including for repositories owned by {% data variables.product.prodname_emus %} and the enterprise.
For personal repositories, a user's own settings determine whether paid usage is enabled, unless the user is managed by an enterprise (EMU). In that case, the enterprise’s settings apply.
For more information about the billing hierarchy between enterprises and organizations, see [AUTOTITLE](/billing/managing-the-plan-for-your-github-account/about-billing-for-plans).
## Further reading
* [AUTOTITLE](/billing/managing-billing-for-your-products/viewing-your-product-usage)

1
content/billing/managing-billing-for-your-products/index.md
View File

@@ -14,6 +14,7 @@ children:
- /about-billing-for-github-codespaces
- /about-billing-for-github-copilot
- /about-billing-for-github-packages
- /about-billing-for-github-models
- /managing-billing-for-github-advanced-security
- /managing-billing-for-git-large-file-storage
- /managing-billing-for-github-marketplace-apps

3
content/billing/managing-your-billing/about-the-new-billing-platform.md
View File

@@ -32,6 +32,7 @@ The products shown in the new billing platform are determined by your {% data va
* {% data variables.product.prodname_registry %}
* {% data variables.large_files.product_name_long %}
* {% data variables.product.prodname_marketplace %}
* {% data variables.product.prodname_github_models %}
* {% data variables.product.prodname_sponsors %}
{% endif %}
@@ -42,6 +43,7 @@ The products shown in the new billing platform are determined by your {% data va
* {% data variables.product.prodname_GH_cs_and_sp %} (only available with {% data variables.product.prodname_team %})
* {% data variables.product.prodname_github_codespaces %}
* {% data variables.product.prodname_copilot %}
* {% data variables.product.prodname_github_models %}
* {% data variables.product.prodname_registry %}
* {% data variables.large_files.product_name_long %}
@@ -54,6 +56,7 @@ The products shown in the new billing platform are determined by your {% data va
* {% data variables.product.prodname_github_codespaces %}
* {% data variables.product.prodname_copilot %}
* {% data variables.product.prodname_enterprise %}
* {% data variables.product.prodname_github_models %}
* {% data variables.product.prodname_registry %}
* {% data variables.large_files.product_name_long %}

1
content/github-models/github-models-at-scale/index.md
View File

@@ -6,4 +6,5 @@ versions:
children:
- /use-models-at-scale
- /manage-models-at-scale
- /set-up-custom-model-integration-models-byok
---

2
content/github-models/github-models-at-scale/manage-models-at-scale.md
View File

@@ -47,4 +47,6 @@ While {% data variables.product.prodname_github_models %} for organizations and
{% data reusables.models.enterprise-change-models-settings %}
{% data reusables.models.enable-select-models-intro %}
{% data reusables.models.enable-select-models-in-org %}

62
content/github-models/github-models-at-scale/set-up-custom-model-integration-models-byok.md Normal file
View File

@@ -0,0 +1,62 @@
---
title: Using your own API keys in GitHub Models
shortTitle: Use custom models
intro: Learn how to integrate your preferred custom models with {% data variables.product.prodname_github_models %} by using your own LLM API keys.
versions:
feature: github-models
permissions: Organization owners can add custom models to {% data variables.product.prodname_github_models %} for their organization
topics:
- Enterprise
allowTitleToDifferFromFilename: true
---
{% data reusables.models.byok-preview-note %}
You can bring your own API keys (BYOK) to {% data variables.product.prodname_github_models %}, and enable teams to use your preferred large language model (LLM) providers across tools like Prompts, Playground, and Models in Actions. For more information about {% data variables.product.prodname_github_models %}, see [AUTOTITLE](/github-models/about-github-models).
To learn about billing and pricing, see [AUTOTITLE](/billing/managing-billing-for-your-products/about-billing-for-github-models).
>[!NOTE] Model support is currently limited to OpenAI and AzureAI.
## Why bring your own API keys?
As an organization owner, you may have specific requirements for governance, data security, and compliance. By setting up your own API keys, you can:
* **Governance and compliance:** Choose LLM providers that comply with your organization's policies and regulatory requirements.
* **Cost management:** Align with your existing payment methods, contracts, credits, or negotiated rates, and avoid usage overages.
* **Visibility and control:** Manage which models your team can access, and monitor usage through your provider's existing dashboards and billing.
* **Flexibility:** Support custom or specialized models that your organization already uses.
## Setting up your API keys to add custom models
> [!IMPORTANT] We highly recommend adhering to the principle of least privilege by assigning only the minimum necessary scopes to your API keys.
You must first add the relevant API keys for the organization. After that, your can specify or enable the custom models you wish to make available to users.
### Adding API keys
{% data reusables.profile.access_org %}
{% data reusables.profile.org_settings %}
{% data reusables.organizations.custom-models %}
1. Click **Add custom key**.
1. In the "Add a custom key" dialog, provide details about your key. **Name** and **Key** are compulsory fields.
1. Click **Save**.
## Enabling custom models
{% data reusables.profile.access_org %}
{% data reusables.profile.org_settings %}
{% data reusables.organizations.models-development %}
1. Under "Models permissions", select **All publishers** to enable models added by custom keys.
* If this option isn't available, you need to allow the use of the model in the organization. See [AUTOTITLE](/github-models/github-models-at-scale/manage-models-at-scale#controlling-model-usage-in-your-organization).
1. Optionally, select **Only select models** to create a custom list of enabled or disabled models. This allows you to control which models are available to your organization.
## Next steps
Now that you've enabled your custom models in {% data variables.product.prodname_github_models %}, learn how to:
* Experiment with your custom model in the playground. See [Experimenting with AI models in the playground](/github-models/use-github-models/prototyping-with-ai-models#experimenting-with-ai-models-in-the-playground).
* Store prompts on {% data variables.product.github %} so that you can iterate to fine-tune your prompts, and share them with stakeholders. See [AUTOTITLE](/github-models/use-github-models/storing-prompts-in-github-repositories).
* Launch your AI application. See [Going to production](/github-models/use-github-models/prototyping-with-ai-models#going-to-production)

3
content/github-models/github-models-at-scale/use-models-at-scale.md
View File

@@ -20,12 +20,13 @@ allowTitleToDifferFromFilename: true
* **Centralized model management:** Control which AI models and providers are available to developers across your organization.
* **AI development at speed:** Quickly prototype, evaluate, and optimize prompts and models.
* **API access:** Use the {% data variables.product.prodname_github_models %} REST API to automate and integrate with enterprise workflows.
* **Custom model integration:** Bring your own LLM API keys to connect external or custom models, giving your organization greater flexibility and control over which models are available in {% data variables.product.prodname_github_models %}, whilst keeping aligned with your existing payment methods, credits, and providers.
* **Governance and compliance controls:** Enforce your organization's standards and monitor model usage.
* **Cost optimization:** Avoid unexpected costs from high-priced models.
* **Collaboration:** Share prompts and results using standard {% data variables.product.github %} development practices.
* **Security-focused architecture:** Rest assured that your data remains within {% data variables.product.github %} and Azure and is not shared with model providers.
* **Visual interface:** Allow non-technical team members to contribute alongside developers.
* **API access:** Use the {% data variables.product.prodname_github_models %} REST API to automate and integrate with enterprise workflows.
* **Version control:** All prompt and model changes go through a standard {% data variables.product.github %} commit and pull request flow so you know when and why a prompt changed.
See [AUTOTITLE](/github-models/about-github-models).

4
content/github-models/use-github-models/prototyping-with-ai-models.md
View File

@@ -10,6 +10,8 @@ redirect_from:
If you want to develop a generative AI application, you can use {% data variables.product.prodname_github_models %} to find and experiment with AI models for free. Once you are ready to bring your application to production, you can switch to a token from a paid Azure account. See the [Azure AI](https://aka.ms/azureai/github-models) documentation.
Organization owners can integrate their preferred custom models into {% data variables.product.prodname_github_models %}, by using an organization's own LLM API keys. See [AUTOTITLE](/github-models/github-models-at-scale/set-up-custom-model-integration-models-byok).
See also [AUTOTITLE](/github-models/responsible-use-of-github-models).
## Finding AI models
@@ -415,5 +417,5 @@ These limits are subject to change without notice.
## Leaving feedback
To ask questions and share feedback, see this [GitHub Models discussion post](https://github.com/orgs/community/discussions/159087).
To ask questions and share feedback, see this [GitHub Models discussion post](https://github.com/orgs/community/discussions/159087).
To learn how others are using {% data variables.product.prodname_github_models %}, visit the [GitHub Community discussions for Models](https://github.com/orgs/community/discussions/categories/models).

2
content/organizations/managing-organization-settings/managing-or-restricting-github-models-for-your-organization.md
View File

@@ -24,4 +24,6 @@ allowTitleToDifferFromFilename: true
## Enabling or restricting models in your organization
{% data reusables.models.enable-select-models-intro %}
{% data reusables.models.enable-select-models-in-org %}

2
data/reusables/models/byok-preview-note.md Normal file
View File

@@ -0,0 +1,2 @@
> [!NOTE]
> The ability to bring your own keys (BYOK) to use custom models with {% data variables.product.prodname_github_models %} for organizations on {% data variables.product.github %} is in {% data variables.release-phases.public_preview %} and subject to change.

4
data/reusables/models/enable-select-models-in-org.md
View File

@@ -1,8 +1,6 @@
You can choose to enable or disable {% data variables.product.prodname_github_models %} for your organization. You can also choose to only allow the use of selected models or model publishers.
{% data reusables.profile.access_org %}
{% data reusables.profile.org_settings %}
{% data reusables.organizations.models %}
{% data reusables.organizations.models-development %}
1. Under "Models", in the "Models in your organization" section, click {% octicon "chevron-down" aria-label="the down arrow" %} beside **Disabled** and select **Enabled** from the dropdown.
> [!NOTE]

3
data/reusables/models/enable-select-models-intro.md Normal file
View File

@@ -0,0 +1,3 @@
You can choose to enable or disable {% data variables.product.prodname_github_models %} for your organization. You can also choose to only allow the use of selected models or model publishers. For more information, see the instructions below.
You can also integrate your preferred external LLM models by bringing your own keys (BYOK) to {% data variables.product.prodname_github_models %}. See [AUTOTITLE](/github-models/github-models-at-scale/set-up-custom-model-integration-models-byok)

1
data/reusables/organizations/custom-models.md Normal file
View File

@@ -0,0 +1 @@
1. In the "Code, planning, and automation" section of the sidebar, click **{% octicon "ai-model" aria-hidden="true" aria-label="ai-model" %} Models** dropdown. Then click **Custom models**. The "Custom Models" page displays the API keys and custom models added to the organization.

1
data/reusables/organizations/models-development.md Normal file
View File

@@ -0,0 +1 @@
1. In the "Code, planning, and automation" section of the sidebar, click the **{% octicon "ai-model" aria-hidden="true" aria-label="ai-model" %} Models** dropdown. Then click **Development**.

1
data/reusables/organizations/models.md
View File

@@ -1 +0,0 @@
1. In the "Code, planning, and automation" section of the sidebar, click **{% octicon "ai-model" aria-hidden="true" aria-label="ai-model" %} Models**.