Skip to content
ModelCap

Inference provider

Google: AI models and API pricing

Every tracked model Google serves through OpenRouter, with Google's own input and output price per 1M tokens, the context window it serves, disclosed quantization and measured endpoint uptime. Rows link to the model page, where every provider's deployment is listed side by side.

Snapshot as of 11 August 2026

Google serves 38 tracked models through 87 endpoints as of 11 August 2026; the highest-ranked is Claude Fable 5 (ModelCap Index #1); mean measured uptime is 98.28%.

Models served
38
On the ModelCap Index
16
Cheapest output price
$0.40/1M

Language models on Google (34)

Ranked models first, in ModelCap Index order. Prices are the lowest Google lists for the model; the endpoint count shows how many deployments carry it.

Language models served by Google, with price, context, quantization and uptime
#ModelModelCap IndexInput $/1MOutput $/1MContextQuantizationUptime (30 min)Endpoints
1Claude Fable 5Anthropic#183.1$10.00$50.001M—87.95%2
2Claude Opus 5Anthropic#279.6$5.00$25.001M—100.0%3
3Gemini 3.1 Pro PreviewGoogle#677.7$1.00$6.001M—98.63%3
4Gemini 3.6 FlashGoogle#876.2$0.75$3.751M—99.86%3
5Claude Sonnet 5Anthropic#1470.0$2.00$10.001M—99.91%3
6Gemini 3.5 Flash LiteGoogle#1766.0$0.15$1.251M—100.0%4
7Gemma 4 26B A4BGoogle#2559.2$0.15$0.60262K—97.25%1Open weights
8Kimi K2 ThinkingMoonshot AI#2757.2$0.60$2.50262K—98.62%1
9Qwen3 235B A22B Instruct 2507Qwen#3153.6$0.22$0.88262K—100.0%1Open weights
10Claude Haiku 4.5Anthropic#3651.0$1.00$5.00200K—100.0%3
11Qwen3 Next 80B A3B InstructQwen#4046.6$0.15$1.20262K—100.0%1Open weights
12Qwen3 Coder 480B A35BQwen#4643.7$0.22$1.80262K——1Open weights
13Qwen3 Next 80B A3B ThinkingQwen#5936.7$0.15$1.20262K——1Open weights
14Llama 4 MaverickMeta#18123.5$0.35$1.15524K—100.0%1
15Llama 4 ScoutMeta#18821.7$0.25$0.701.3M—100.0%1
16Llama 3.3 70B InstructMeta#19320.3$0.72$0.72128K—99.05%2
17Claude Opus 4Anthropic—$15.00$75.00200K—100.0%1
18Claude Opus 4.1Anthropic—$15.00$75.00200K——1
19Claude Opus 4.5Anthropic—$5.00$25.00200K——1
20Claude Opus 4.6Anthropic—$5.00$25.001M—100.0%2
21Claude Opus 4.7Anthropic—$5.00$25.001M—100.0%3
22Claude Opus 4.8Anthropic—$5.00$25.001M—100.0%3
23Claude Sonnet 4Anthropic—$3.00$15.001M—100.0%3
24Claude Sonnet 4.5Anthropic—$3.00$15.001M—100.0%2
25Claude Sonnet 4.6Anthropic—$3.00$15.001M—100.0%2
26Gemini 2.5 FlashGoogle—$0.30$2.501M—95.17%4
27Gemini 2.5 Flash LiteGoogle—$0.10$0.401M—99.87%2
28Gemini 2.5 ProGoogle—$0.625$5.001M—95.32%5
29Gemini 2.5 Pro Preview 05-06Google—$0.625$5.001M—95.60%5
30Gemini 2.5 Pro Preview 06-05Google—$0.625$5.001M—95.32%5
31Gemini 3 Flash PreviewGoogle—$0.25$1.501M—99.50%3
32Gemini 3.1 Flash LiteGoogle—$0.125$0.751M—98.70%4
33Gemini 3.5 FlashGoogle—$0.75$4.501M—99.80%4
34GLM 4.7Z.ai—$0.60$2.20200K—99.93%1Open weights

Image, audio and video models on Google (4)

Media models are ranked within their own catalogues; token prices do not apply.

Image, audio and video models served by Google
#ModelContextQuantizationUptime (30 min)Endpoints
1Nano Banana Pro (Gemini 3 Pro Image)Google66K—100.0%1
2Nano Banana 2 (Gemini 3.1 Flash Image)Google131K—99.86%1
3Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)Google66K—99.78%1
4Nano Banana (Gemini 2.5 Flash Image)Google33K—98.64%2

Google on ModelCap: common questions

How many AI models does Google serve?

38 tracked models as of 11 August 2026: 34 language models (16 with a public ModelCap Index position) and 4 image, audio or video models, across 87 live endpoints observed through OpenRouter.

What is the best model available on Google?

Claude Fable 5 from Anthropic is the highest-ranked language model Google serves, at ModelCap Index position #1 and $50.00 per 1M output tokens on Google. The table lists every model in Index order.

What is the cheapest model on Google?

Gemini 2.5 Flash Lite at $0.40 per 1M output tokens ($0.10 per 1M input), the lowest positive language-token price Google lists. Free-tier endpoints, where present, show as $0 and are excluded from that comparison.

Which Google model has the largest context window?

Llama 4 Scout, served with a 1.3M-token context window on Google. Providers sometimes serve less than the model's published maximum, so the figure here is the provider's own.

How reliable is Google?

Mean measured uptime across Google endpoints is 98.28% over the last 30 minutes, as reported by OpenRouter at the time of the snapshot. Per-model uptime is in the table; it is a live operational signal, not a service-level guarantee.

Where do the Google prices come from?

From the OpenRouter catalogue entry for each Google endpoint, in USD per 1M tokens, re-observed on every ModelCap refresh. Where Google exposes several deployments of one model at different prices, the table shows the lowest and notes the count; the model page lists each deployment.

How current is this page?

It re-renders within a minute of every dataset refresh; this snapshot was published 11 August 2026. Offers retained from a failed refresh are marked stale in the data and are not counted here.

Explore further