# Gemini 4 Neon

> Gemini 4 Neon is Google's fast frontier model between Carbon and Hydrogen for software engineering, knowledge work and science. It reads text, images and video, returns text, and has a 1,048,576-token context window.

**ID**: `google/gemini-4-neon`  
**Creator**: Google  
**Category**: text  
**Context**: 1M tokens  
**Pricing**: $3.00 / $15.00 per 1M · billed by your provider (BYOK)  
**Released**: 2026-11-18  
**Web page**: https://anyrouter.dev/model/google/gemini-4-neon

**Input modalities**: text, image, video  
**Output modalities**: text  

**Tokenizer**: Gemini

**Capabilities**: chat, coding, vision, reasoning

## Usage

**Endpoint**: `POST https://anyrouter.dev/api/v1/chat/completions`

```bash
curl https://anyrouter.dev/api/v1/chat/completions \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "google/gemini-4-neon",
  "messages": [
    {
      "role": "user",
      "content": "Say hi in 3 words."
    }
  ]
}'
```

## Providers

| Provider | Input | Output |
| --- | --- | --- |
| google-aistudio-byok | $3.00/M (list) | $15.00/M (list) |
| openrouter-byok | $3.00/M (list) | $15.00/M (list) |
| vercel-byok | $3.00/M (list) | $15.00/M (list) |
| respan-byok | $3.00/M (list) | $15.00/M (list) |
| infron-byok | $3.00/M (list) | $15.00/M (list) |
| experiential-byok | $3.00/M (list) | $15.00/M (list) |

BYOK routes are billed by your own provider key — AnyRouter charges $0. Rates marked (list) are the provider's published list price.

## Model card


### Strengths [2,3]

- Google reports it is 3x faster than Gemini 4 Carbon on Glean evaluations
- Listed at $3 per 1M input and $15 per 1M output tokens, with 32% lower cost per task than Carbon at high reasoning effort (Google-reported)
- Google reports an AA Index of 61 versus Claude Opus 5.5 at 59 and GPT-6.2 Sol at 57, 39% on SWE-sweep, and 71.3% on HLE-Diamond

### Known limits [1]

- The speed and benchmark figures come from Google's own evaluation and have not been independently reproduced here
- Audio and PDF input are listed by Google but are not yet exposed through AnyRouter's chat surface

### Use cases [1,2]

- Complex software engineering and terminal coding
- Knowledge work and long-context document analysis
- Scientific reasoning and agentic workflows

### API notes [1]

- 1M-token context with up to 64K output tokens per response, extendable through continuation
- Thinking depth can be controlled with low, medium, high and xhigh reasoning effort

### Links

- announcement: https://x.com/GoogleDeepMind/status/2123167884411077030
- docs: https://ai.google.dev/gemini-api/docs/models
- pricing: https://ai.google.dev/gemini-api/docs/pricing
- Previous version: https://anyrouter.dev/model/google/gemini-4-carbon

### Sources

1. [Gemini API models](https://ai.google.dev/gemini-api/docs/models) (accessed 2026-11-19)
2. [Gemini 4 Neon launch announcement](https://x.com/GoogleDeepMind/status/2123167884411077030) (accessed 2026-11-19)
3. [Gemini API pricing](https://ai.google.dev/gemini-api/docs/pricing) (accessed 2026-11-19)

## Source

- Upstream docs: https://ai.google.dev/gemini-api/docs/models
