Select Gemini 4 Neon directly in the model picker in Cursor Chat or Agent. Cursor lists it as a third-party model and deducts usage charges from the prepaid balance on individual plans under The X Subscription or from the Other Models usage pool on eligible legacy plans. Model selection · Neon pricing
Suitable tasks. Day-to-day coding with Agent and tasks that require repeated tool calls, such as reading files and running commands.
Unsuitable tasks. Cursor recommends Gemini 4 Carbon for the most difficult tasks.
Applicable model version. Gemini 4 Neon as listed by Cursor, with the model ID gemini-4-neon.
Applicable clients, agents, or APIs. Cursor Chat and Agent.
Recommended reasoning levels and parameters. Cursor's Neon model page gives no available reasoning levels or default value. Cursor model documentation
| Setting | What to select in Cursor |
|---|---|
| Model | Open the model picker in the Chat or Agent panel and select Gemini 4 Neon. Cursor keeps your selection until you switch models manually. |
| Team access | Administrators can view and manage model access under Team Settings > Models. |
| The X Subscription for individuals | Neon is available on Premium, Super, and Ultra. Cursor charges the prepaid balance starting with the first request. Free and Plus lack access to third-party models. |
| Legacy Cursor plans | Pro, Pro+, Ultra, and legacy Teams have an Other Models usage pool. |
Model selection and permissions · Subscription rules for third-party models · Legacy usage pools
Open the Chat or Agent panel in Cursor, select Gemini 4 Neon in the model picker, and confirm the active model name before sending a request. Model selection instructions
Use a small repository task to check which tools Agent actually calls. For example, ask Agent to find the application's entry point and list the files it read. In Agent, Neon can search for, read, and edit files, run Shell commands, browse the web, generate images, ask questions, and read rules. Neon tool list
Check the charge for that request in the usage dashboard, then estimate what a multi-turn task would cost on your current plan. On legacy plans, check the Other Models usage pool. Cursor usage dashboard · Legacy usage documentation
Cursor's model page lists a 200K context window and a 1M maximum context. The table shows Cursor's published prices at the time of collection in US dollars per million tokens. Cursor model and pricing
| Input | Cache write | Cache read | Output |
|---|---|---|---|
| $3.00 | No separate rate listed | $0.15 | $15.00 |
Cursor charges an additional $0.25 per million tokens for qualifying third-party requests on Teams and Enterprise, including input, output, and cached tokens. When you select Neon directly, its cache read rate of $0.15 per million tokens exceeds the exemption threshold. The surcharge also applies to Other Models usage on legacy Teams plans. Cursor Token Rate
The models available to an account depend on its plan, team permissions, and provider regional restrictions. Model availability
Cursor lists 1M as the maximum context. The context window you can use depends on the modes your client and plan offer. Model specifications · Cursor Max Mode rules
The settings and prices above apply to Cursor. Check which reasoning options Neon shows in your account.
At the time of collection, this note was checked against Cursor's Neon model page, model selection help, and plan documentation. The client steps are for readers to try. Actual tool calls and billing charges still need to be verified in a Cursor account.
Gemini 4 Neon