Gemini 4 Neon

LiveGoogle
Model IDgemini-4-neon

Gemini 4 Neon by Google: 1M context, 1M max output, with Vision, Function Calling, Reasoning, Prompt Caching, Response Schema, Audio, Video, File Input, Streaming, Web Search. Via SoleAPI, Input (prompt) costs 10.5 Credits / M tokens, 47% below list price. 7-day uptime 37.0%. Compatible with Anthropic Messages / Gemini generateContent: swap the base URL and you are connected.

Context window
1Mtokens
1,048,576
Max output
1Mtokens
1,048,576
7-day uptime
37.0%
108 probes · 68 failed
Measured throughput
66.2tokens / s
Real traffic, last 7 days · output side

Capabilities

Listed exactly as the catalog flags them, including what this model does not support.

10 of 15 supported

Vision

Natively multimodal: captioning, classification, visual Q&A, chart and screenshot understanding. Images are converted to tokens at the input rate.

Function Calling

Connect the model to external tools and APIs: it decides when to call which function and with what arguments, your code runs it. Force or forbid calls with tool_choice.

Reasoning

The model thinks before answering, which markedly improves multi-step planning and reasoning for coding, math and data analysis. Thinking tokens bill at the output rate.

Prompt Caching

Repeated prefixes (system prompts, long documents) hit the cache: reads cost far less than input, writes are priced by retention time.

Response Schema

Constrain output to a JSON Schema you provide for predictable, type-safe results without parsing free text.

Audio

Audio input or speech output.

Video

Video input understanding or video generation.

File Input

Upload PDFs and other documents directly; the model reads the whole document with native vision.

Streaming

Tokens arrive over SSE as they are generated; the first frame is forwarded immediately with no buffering. Claude Code, Codex and similar tools depend on this.

Web Search

Vendor-hosted live web retrieval; the model decides when to search and cites sources.

RealtimeNot supported

Bidirectional low-latency voice / text sessions over WebSocket.

SessionNot supported

Server-side multi-turn context (e.g. previous_response_id in Responses) so you need not resend history.

MediaNot supported

Image / video / music generation endpoints.

Multipart FormNot supported

multipart/form-data file-upload requests.

Binary ResponseNot supported

Endpoints that return raw audio / image bytes.

Reliability & performance

SoleAPI probes this model through the real supply chain on a schedule — measured, not self-reported.

Last probe: 11/22/2026, 1:09:57 PM
37.0%108 probes in 7 days, 68 failed · automatic failover to backup channels
✕ 11/21/2026, 4:35:22 PM✕ 11/21/2026, 4:56:43 PM✕ 11/21/2026, 5:17:51 PM✕ 11/21/2026, 5:38:42 PM✕ 11/21/2026, 5:59:01 PM✕ 11/21/2026, 6:20:27 PM✕ 11/21/2026, 6:41:51 PM✕ 11/21/2026, 7:02:27 PM✕ 11/21/2026, 7:23:51 PM✕ 11/21/2026, 7:45:00 PM✕ 11/21/2026, 8:06:22 PM✕ 11/21/2026, 8:27:48 PM✕ 11/21/2026, 8:48:39 PM✕ 11/21/2026, 9:09:08 PM✕ 11/21/2026, 9:30:21 PM✕ 11/21/2026, 9:52:51 PM✕ 11/21/2026, 10:13:26 PM✕ 11/21/2026, 10:34:51 PM✕ 11/21/2026, 10:55:25 PM✕ 11/21/2026, 11:16:51 PM✕ 11/21/2026, 11:39:11 PM✕ 11/22/2026, 12:00:21 AM✕ 11/22/2026, 12:21:52 AM✕ 11/22/2026, 12:43:21 AM✕ 11/22/2026, 1:04:51 AM✕ 11/22/2026, 1:26:09 AM✕ 11/22/2026, 1:46:24 AM✕ 11/22/2026, 2:07:51 AM✕ 11/22/2026, 2:28:56 AM✕ 11/22/2026, 2:49:27 AM✕ 11/22/2026, 3:09:54 AM✕ 11/22/2026, 3:30:24 AM✕ 11/22/2026, 3:50:54 AM✕ 11/22/2026, 4:11:24 AM✕ 11/22/2026, 4:32:10 AM✕ 11/22/2026, 4:54:43 AM✕ 11/22/2026, 5:15:56 AM✕ 11/22/2026, 5:36:51 AM✕ 11/22/2026, 5:57:55 AM✕ 11/22/2026, 6:19:09 AM✕ 11/22/2026, 6:39:32 AM✕ 11/22/2026, 7:01:02 AM4.4s3.9s2.8s3.0s4.4s4.5s6.3s13.7s5.2s4.4s4.5s4.1s8.3s2.6s✕ 11/22/2026, 12:08:23 PM✕ 11/22/2026, 12:28:54 PM✕ 11/22/2026, 12:49:28 PM✕ 11/22/2026, 1:09:57 PM
11/21/2026, 4:35:22 PMnow
successful probe, bar height = end-to-end latency (0–4 s)failed probegrid 1 s

Latency & throughput

Latest latency2.6s
Avg. of last 605.2s
P95 of last 6013.7s
Output throughput66.2t/s
ProtocolsAnthropic Messages / Gemini generateContent

Protocols & code samples

Change the base URL to SoleAPI, keep everything else. Native protocols pass straight through; “adapted” ones are translated by the gateway.

Gemini CLI setup guide

Anthropic SDK or any Messages-protocol client: point base_url at https://api.soleapi.com and use a SoleAPI key.

curl https://api.soleapi.com/v1/messages \
  -H "x-api-key: $SOLEAPI_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "gemini-4-neon",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Explain prompt caching in one sentence"}]
  }'

Full price list

List price → SoleAPI price per channel, every pricing dimension, nothing hidden.

1 Credit ≈ $0.151 · rate updated daily
ItemUnitGoogle list priceOfficial Gemini Group ×3.547% off
Input (prompt)per 1M tokens$3.0010.5 Credits
Output (completion)per 1M tokens$15.0052.5 Credits
Cache readper 1M tokens$0.150.525 Credits
Cache write (general)per 1M tokens$3.7513.13 Credits
Web search (general)per 1K calls$14.0049 Credits

“Off” is computed at 1 Credit = ¥1 and today’s rate: 1 − channel multiplier × (1 Credit in the list currency). Priority: when request body field service_tier matches ^priority$, billed at Cache read ×1.8、Cache write (general) ×1.8、Output (completion) ×1.8、Input (prompt) ×1.8、other items ×1.8; stacks with any channel.

FAQ

How much does the Gemini 4 Neon API cost?

Via SoleAPI, Input (prompt) costs 10.5 Credits / M tokens, with 1 Credit = ¥1. That is 47% below the Google list price.

What are the context window and output limit of Gemini 4 Neon?

Context window 1M tokens, up to 1M tokens per response.

Does Gemini 4 Neon support function calling and structured outputs?

Function calling with tool_choice control is supported; structured outputs (JSON Schema) are supported.

How do I use Gemini 4 Neon in Claude Code?

Set ANTHROPIC_BASE_URL=https://api.soleapi.com, ANTHROPIC_API_KEY to your SoleAPI key and ANTHROPIC_MODEL=gemini-4-neon, then run claude. No config files to edit.

How is the 37.0% uptime measured?

SoleAPI sends probe requests to this model through the real supply chain on a schedule and divides successful probes by total probes over the last 7 days. When one channel fails the gateway fails over to a backup, so the figure reflects what end users actually see.

One API key for Gemini 4 Neon and 30+ models

Free credits on sign-up, pay as you go, no monthly fee.

Get an API Key