Cursor on Vana API: GPT pay-as-you-go pricing from $0.1278

§ What does Cursor need to connect to Vana API?

Cursor is one of the clients listed for the OpenAI format on Vana API. That means you can point Cursor at the single Vana API address, https://api.vanaapi.com, and use models such as GPT-5.6 Sol, GPT-6 Astra, GPT-6.1 Sol, Grok 4.7, DeepSeek V4 Flash, DeepSeek V4 Pro, GLM-5.3, Kimi K3, Qwen3.7 Plus, Qwen3.8 Max, MiniMax M3 and MiMo V2.6 Pro. The endpoint for these OpenAI-format requests is /v1/chat/completions.

The setup is deliberately small. You create a key in the Vana API dashboard, add the base address to Cursor, and choose the model you want. The address itself answers API requests. Keys, balance and the usage log are handled in the dashboard on the main site, https://vanaapi.com. This separation keeps the configuration focused: one address for requests, one dashboard for account details.

Vana API is built around one address and three protocols. Cursor uses the OpenAI-compatible protocol, so the same address can serve several kinds of clients without asking you to maintain a separate endpoint for each one. You can also use an OpenAI SDK or cURL with the OpenAI format, while GPT models additionally answer on /v1/responses, which is the route used by Codex.

The practical question is not whether Cursor is supported. It is which model and billing approach fit your work. Vana API uses a prepaid balance, charges per real token, and has no monthly fee. Every call appears in the usage log in the dashboard, so you can check what was spent and compare it with the token prices below.

§ Which models can I select in Cursor?

Cursor can use the GPT, Grok, DeepSeek, GLM, Kimi, Qwen, MiniMax and MiMo models listed for the OpenAI format. The model identifiers are explicit, which helps when you are checking a configuration or moving a setting between tools:

• GPT-5.6 Sol: gpt-5.6-sol
• GPT-6 Astra: gpt-6-astra
• GPT-6.1 Sol: gpt-6.1-sol
• Grok 4.7: grok-4.7
• DeepSeek V4 Flash: deepseek-v4-flash
• DeepSeek V4 Pro: deepseek-v4-pro
• GLM-5.3: glm-5.3
• Kimi K3: kimi-k3
• Qwen3.7 Plus: qwen3.7-plus
• Qwen3.8 Max: qwen3.8-max
• MiniMax M3: MiniMax-M3
• MiMo V2.6 Pro: MiMo-V2.6-Pro

Price is shown per million tokens for input and output. For example, GPT-5.6 Sol is $0.3472 / $1.81 on Vana API, compared with an official list price of $4 / $20. GPT-6.1 Sol is $0.1389 / $0.8333, while GPT-6 Astra is $0.6944 / $3.54. These figures let you select a model by cost as well as by the kind of response you expect from Cursor.

Other choices have their own input and output rates. Grok 4.7 is $0.4861 / $1.53. DeepSeek V4 Flash is $0.2778 / $1.04, and DeepSeek V4 Pro is $1.11 / $3.33. GLM-5.3 is $0.5556 / $1.67, Kimi K3 is $1.04 / $5.21, Qwen3.7 Plus is $0.1278 / $0.4167, Qwen3.8 Max is $0.6944 / $2.08, MiniMax M3 is $0.1389 / $0.5556, and MiMo V2.6 Pro is $0.2083 / $0.3472.

Because the rates are listed per million tokens, a small comparison can be useful. For example, say you send 1,000,000 input tokens and receive 100,000 output tokens to GPT-5.6 Sol. Using the listed Vana API rates, that example would cost $0.3472 for input plus $0.181 for output, based on $1.81 per million output tokens. This is an example, not a quotation for a particular workload. The usage log remains the record of what your own calls cost.

§ How do I set up Cursor with one address?

The setup process has three short parts. Create a key in the dashboard, enter https://api.vanaapi.com as the API address in Cursor, and select a model identifier from the list. The OpenAI format uses /v1/chat/completions, so Cursor can send requests through the route intended for OpenAI-format clients. There is no need for a separate relay address for each supported model.

Account information stays on the main Vana API site. You can check your balance, review the usage log and create keys at https://vanaapi.com. The API address does not provide dashboard functions; it answers API requests. Keeping those roles separate makes it easier to confirm that a request is reaching the right endpoint and that a key is active.

For users who prefer not to configure a client manually, the same OpenAI-format address can be used with an OpenAI SDK or cURL. Cursor, Cline, Roo Code, OpenCode and Cherry Studio are also listed for that format. The common factor is the protocol, not the client. Vana API presents one address with three protocols, and Cursor uses the OpenAI-format route.

A sign-up gift of $0.10 of credit is available. It is small, but it gives you a starting balance to test a request before adding funds. Since Vana API charges per real token and has no monthly fee, the balance is a direct way to track consumption rather than a subscription renewal.

§ Should I use prepaid balance or a monthly plan?

The clearest criterion is whether your usage is predictable and concentrated on GPT models. Vana API’s monthly plan is described as GPT up to $4,000 / 30 days: $103.61 for 30 days, with at most $1,200 a week, and it covers GPT models only. If you need a scheduled allowance for GPT work, this is the plan option described by Vana API.

For variable usage or for a mix of models, the prepaid balance is more flexible. You top up, spend against the balance, and see each call in the usage log. There is no monthly fee on that route. If you switch between GPT and models such as DeepSeek V4 Flash or Qwen3.7 Plus, the per-token rates remain visible on the price list.

The referral programme can also affect the value of a balance. You receive 14% of what people you invite pay, and 5% one level further. Earnings can be withdrawn from $13.89. These percentages apply to payments by people you refer, not to token usage, so they are separate from the per-token pricing decision.

Refunds are available for unused balance within 3 days of the top-up, minus what was used and the payment processor's fee. If you are comparing the monthly plan with prepaid balance, this refund window is one of the concrete points to check before you top up.

§ What should I verify before using Cursor with Vana API?

Begin with the protocol. Cursor is an OpenAI-format client, so its requests should use the OpenAI route on the single Vana API address. Check that the model identifier matches the model you selected in Cursor, particularly for names with a version suffix such as gpt-6.1-sol or MiniMax-M3. The same address also supports other protocols, but the client determines which route it uses.

Next, check the key and the balance in the dashboard. Keys are created on https://vanaapi.com, and every call is listed in the usage log. If you are testing cost, start with a model whose rate is easy to locate, such as Qwen3.7 Plus at $0.1278 / $0.4167 or GPT-6.1 Sol at $0.1389 / $0.8333. The log then gives you a direct view of how the listed rates translate into your own use.

Finally, decide whether the GPT monthly plan matches your pattern. The stated limit is $103.61 for 30 days, with at most $1,200 a week, for GPT models only. Otherwise, prepaid balance keeps the commitment tied to actual token consumption. Either way, the setup remains the same: one address, three protocols, minimal setup, and a dashboard that shows the details behind each call.