Qoder BYOK: custom model and Base URL setup (November 2026)
As of 2026-11-10, Qoder can use your own API key (individual plan): on the desktop app go to Settings > Models > Add model and pick OpenAI Compatible or Anthropic Compatible under Custom, then enter the Base URL, an API key and a Model ID; in the CLI, add the model with the wizard on the Custom tab of /model. Per the official release notes, Qoder 0.1.8 (2026-09-05) made personal-plan BYOK accept a custom Base URL. With QCode: pick Anthropic Compatible and enter https://api.qcode.cc/api for Claude, or pick OpenAI Compatible (Chat Completions API) and enter https://api.qcode.cc/openai/v1 for GPT and Chinese models, with your cr_ key.
Updated 2026-11-10
Point this tool at the QCode endpoint — one key for Claude, GPT and Chinese models.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
Four points from the official sources
OpenAI and Anthropic compatible
The official page lists two custom providers, OpenAI Compatible and Anthropic Compatible; after OpenAI Compatible you choose Chat Completions API or Responses API.
Custom Base URL since 2026-09-05
Qoder 0.1.8 release note: personal plan BYOK now supports custom Base URLs for any OpenAI- or Anthropic-compatible model service. As of 2026-11-10 the latest desktop version is 0.5.7 (2026-11-09).
Desktop entry point
Settings, then Models in the left navigation, then Add model; under Custom pick the provider type, enter the Base URL, API key and at least one Model ID, and click Validate and add model.
Does not use Qoder Credits
Official FAQ: custom model fees are billed directly to your provider API account and do not use Qoder Credits; Repo Wiki uses a fixed model and still consumes Credits. Image and video generation are not available with BYOK models.
Can Qoder use your own API key?
As of 2026-11-10, yes. The official custom models page says: "Qoder supports accessing third-party provider model resources via API keys." The CLI page says "Available edition: Individual plan", and the pricing page lists "Bring Your Own Key (BYOK)" in the Free column of the individual plans. The desktop app, the CLI and QoderWork differ. Desktop: Settings > Models > Add model, then a custom endpoint with a Base URL, API key and Model ID. CLI: enter /model, switch to the Custom tab and select Add custom model... to follow the wizard. QoderWork: its official models page lists Premium, Advanced and Standard plus Qwen3.7-Max and Qwen3.7-Plus, with no custom model entry as of 2026-11-10; the Qoder download page says QoderWork will officially cease operations on 2027-09-05, remains available to download and use until then, and recommends moving to Qoder Desktop in advance, which has built-in data import tools. The global site qoder.com and the China site Qoder CN use the same custom steps; only the preset provider lists differ, and the Qoder CN list has no OpenAI, Google or OpenRouter.
2026-09-05: Qoder 0.1.8 adds custom Base URLs to personal-plan BYOK
The Qoder 0.1.8 release note (2026-09-05) is titled "Custom Base URLs for BYOK" and says personal plan BYOK now supports custom Base URLs for any OpenAI- or Anthropic-compatible model service. On the CLI side, the 1.1.50 release note (2026-09-11) says personal edition users can configure and manage BYOK models with custom URL endpoints from the Custom page in /model, and on 2026-09-12 the official Qoder account on X wrote that starting with v1.1.50, Qoder CLI supports custom model integration for all individual users: you enter your Base URL, Model ID and API key to connect. An earlier CLI 0.2.5 note (2026-04-29) already lists "BYOK Custom URL". As of 2026-11-10 the latest desktop version is 0.5.7 and the CLI is 1.1.86, both released on 2026-11-09.
Timeline
Qoder IDE 0.16.0 (the IDE is a separate product from the desktop app) launches the Community Edition; the release note says the Community Edition has opened BYOK, with mainstream Chinese MaaS services such as Alibaba Cloud, Kimi, DeepSeek, Z.ai and MiniMax.
Qoder 0.1.8: personal plan BYOK supports custom Base URLs for any OpenAI- or Anthropic-compatible model service.
Qoder CLI 1.1.50: personal edition users can configure BYOK models with custom URL endpoints from the Custom page in /model.
What Qoder documents, and what it leaves open
Documented by Qoder
① Under Custom on the desktop app there are OpenAI Compatible and Anthropic Compatible; after OpenAI Compatible you select Chat Completions API or Responses API. ② You enter the Base URL, API key and at least one Model ID; Add Model ID adds more models for the same endpoint. ③ In the official words: enter the exact Base URL supplied by the model service; Qoder displays https://api.example.com/v1 as an example format. ④ Validate and add model checks the connection and saves the model only after validation succeeds. ⑤ Fees are billed directly to your provider API account and do not use Qoder Credits. ⑥ Image and video generation are not available with BYOK models. ⑦ The CLI page says not to configure BYOK manually in settings.json and to use the Custom wizard in /model. ⑧ If adding a model fails, the official checks are: whether the Base URL is complete and the protocol and API type match, whether the key is correct, whether the Model ID is correct, whether the balance is sufficient, and whether the network is normal. ⑨ The pricing page lists "Bring Your Own Key (BYOK)" in the Free column of the individual plans.
What the official pages leave open
As of 2026-11-10: ① The desktop custom models page itself does not state an edition; the CLI page says "Available edition: Individual plan" and the desktop release note says "Personal plan BYOK". ② Qoder IDE is a separate product in the same family; its custom models page lists preset providers only and says it applies to users with Individuals plans, while custom Base URLs belong to the new desktop app and the CLI. ③ The CLI page does not list the wizard fields one by one and says the provider catalog is whatever the wizard shows live; the custom URL endpoint wording comes from the release note and the official X post. ④ The docs do not say, protocol by protocol, which path Qoder appends to the Base URL, and give a single example format, https://api.example.com/v1; the result of Validate and add model is the practical test. ⑤ Some community users report that custom models disappear after a trial ends; the official pricing page lists BYOK in the Free column, and the CLI page's troubleshooting step is to confirm that BYOK is available for the current account and plan.
Desktop and CLI entry points
Desktop: Settings > Models
Click Add model, pick OpenAI Compatible or Anthropic Compatible under Custom, enter the Base URL, API key and Model ID, then click Validate and add model. The China site Qoder CN uses the same steps.
CLI: /model > Custom
Enter /model, switch to the Custom tab, select Add custom model... and press Enter, follow the wizard through Provider, model type and model, and enter the API key; after validation it saves automatically, and d deletes a selected model.
How to use your own API key in Qoder and set the Base URL
① Install the desktop app from the Qoder download page (macOS 12+, Windows 10+ or Linux) and sign in; for the CLI use curl -fsSL https://qoder.com/install | bash (Windows PowerShell: irm https://qoder.com/install.ps1 | iex). ② Desktop: Settings > Models in the left navigation > Add model > open Provider, and under Custom pick OpenAI Compatible or Anthropic Compatible; for OpenAI Compatible also pick Chat Completions API or Responses API. ③ Enter the Base URL, API key and Model ID, set the display name, context window, vision and thinking options to match the model, click Validate and add model, then pick the model in the chat. ④ CLI: /model > Custom > Add custom model..., follow the wizard and do not edit settings.json by hand. ⑤ Enter the exact Base URL your model service gives you; Qoder shows https://api.example.com/v1 as the example format. The docs do not list the path Qoder appends for each protocol, so if validation fails, check the address against your provider's docs first.
Running Qoder with a QCode API key
As of 2026-11-10, one QCode key serves both Anthropic and OpenAI compatible endpoints, billed per token with per-model rates on /models. For Claude pick Anthropic Compatible, enter https://api.qcode.cc/api as the Base URL and a Model ID such as claude-sonnet-5-5 or claude-opus-5-5. For GPT pick OpenAI Compatible with Chat Completions API, enter https://api.qcode.cc/openai/v1 and a Model ID such as gpt-6.1-sol; to use the Responses API pick it and enter the same https://api.qcode.cc/openai/v1 (on QCode /openai/responses and /openai/v1/responses are equivalent), with a GPT model. GLM, Kimi, DeepSeek and Qwen run on Anthropic Compatible or on OpenAI Compatible with Chat Completions API, for example deepseek-v4.1-flash, glm-5.3, kimi-k3 and qwen3.8-max; model IDs follow /models. Leave off any trailing slash, click Validate and add model, and go by Qoder's validation result; if Anthropic Compatible does not validate, change the address to https://api.qcode.cc/api/v1 and validate again. Three steps: ① register and top up (Alipay and other major payment methods); ② buy a Personal Plan or a Pay-as-you-go Pack and copy your API key, which starts with cr_; ③ enter it in Qoder as above; the QCode docs article https://docs.qcode.cc/en/docs/ide/qoder walks through the same setup. Not ready for a monthly plan? Pay-as-you-go Packs start at $4 for a one-time quota you use within 7 days.
Frequently asked questions
Can Qoder use my own API key?
Yes. The official custom models page says Qoder supports accessing third-party provider model resources via API keys; the desktop entry is Settings > Models > Add model and the CLI entry is the Custom tab of /model. The CLI page gives the edition as Individual plan. On the desktop app, custom endpoints come in two types, OpenAI Compatible and Anthropic Compatible. The QCode values are above.
How do I set the Base URL in Qoder?
Under Add model, pick a custom endpoint and fill in the Base URL; the official page asks for the exact Base URL supplied by the model service and shows https://api.example.com/v1 as an example format. For QCode, Anthropic Compatible takes https://api.qcode.cc/api and OpenAI Compatible takes https://api.qcode.cc/openai/v1, with no trailing slash; if Anthropic Compatible does not validate, try https://api.qcode.cc/api/v1. If validation fails, the official advice is to check first that the Base URL is complete and the protocol and API type match the endpoint, then the key, Model ID, balance and network.
Does Qoder still use Credits when I bring my own key?
The official FAQ answers: custom model fees are billed directly to your provider API account and do not use Qoder Credits. The exception sits in the same answer: Repo Wiki uses a fixed model and consumes Qoder Credits, with a prompt when generating. With a QCode key, usage is billed per token, with per-model rates on /models.
How do I add a custom model in Qoder CLI?
In the CLI enter /model, switch to the Custom tab, select Add custom model... and press Enter, then follow the wizard through Provider, model type and model and enter the API key and any other information it asks for; after validation the configuration saves automatically and you can switch to it at once. The official page says not to configure BYOK manually in settings.json. If the Custom tab is missing, confirm that BYOK is available for your account and plan.
Can I add a custom model in QoderWork?
As of 2026-11-10, QoderWork's official models page lists Premium, Advanced and Standard plus Qwen3.7-Max and Qwen3.7-Plus, with no custom model entry. The Qoder download page says QoderWork will officially cease operations on 2027-09-05, remains available to download and use until then, and that your Qoder account and credits stay usable with other Qoder products; it recommends moving to Qoder Desktop in advance, whose built-in data import tools bring over task history, skills and memories. To bring your own model, add it in the Qoder desktop app or CLI as described above.
Can I use my own API key on Qoder's Free plan?
Yes. As of 2026-11-10, Qoder's pricing page lists "Bring Your Own Key (BYOK)" in the Free column of the individual plans, and on 2026-09-12 the official Qoder account on X wrote that starting with v1.1.50, Qoder CLI supports custom model integration for all individual users. If the Custom tab or add entry is missing, the CLI page says to confirm that BYOK is available for the current account and plan. Separately, the official notice (published 2026-09-04) says the Lite model tier was removed from the model selector in all Qoder products on 2026-09-18 at 14:00 Singapore time (UTC+8); that change concerns a model tier, and the pricing page still lists a Lite plan.
Sources
Qoder: the custom models pages on docs.qoder.com (Chinese, English, Japanese), the CLI custom models page, the model selector page, the pricing page, the QoderWork models page and the Lite tier retirement notice, the custom models page on docs.qoder.cn, the desktop, CLI and Qoder IDE release notes on changelog.qoder.com, the qoder.com download page and the official Qoder X post of 2026-09-12, all read on 2026-11-10. QCode: qcode.cc/models, qcode.cc/pricing, qcode.cc/pay-as-you-go and the Endpoints & API Paths page on docs.qcode.cc, retrieved on 2026-11-10.
Run Qoder on your own key
One QCode API key in Qoder: Claude through Anthropic Compatible, GPT and Chinese models through OpenAI Compatible, billed per token; per-model rates are on /models.
Related reading
Pay-as-you-go Packs
One-time quota packs valid for 7 days, no subscription; extra packs add to the same key
Cursor custom API key and Base URL
Where keys go in Settings → Models, which slot can change the Base URL, why Tab skips your key, and the QCode setup.
Trae custom model setup guide
How to fill in a Trae custom model: API format, request URL, model ID and API key, plus the Auto mode limit and common errors.
Crypto payment guide
Nine coins, 56 networks, zero platform fee for AI coding API credits.
Statements on this page were checked on 2026-11-10 against the official Qoder docs and release notes; the official pages prevail. Qoder ships frequently, so menu names and options follow the version you install. QCode model availability follows /models.