Skip to content

Google · Chat

Gemini 4 Hydrogen

Google 🇺🇸ChatGemini familyOne of its sources says the weights aren't publicly available.
From Modelrunner AI's listing:
Write and fix code, run tool-calling agents and answer hard questions from a text prompt with Google's speed-focused Gemini 4 thinking model and its 1M-token context.
llminfo IQ
146.6
Maker's price
$3.00 blended / 1M
Max context
1M
Listings
47

Evidence and sources

llminfo IQ 146.6, ranked among chat models. From 107 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 125.7, 95% interval 115–137. From 6 benchmark families across 5 sources. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedGood value (65.1). Cheaper per task than its benchmark results predict. From 16 benchmarks with measured task costs. Plenty of data behind it.How Value for money worksYap score: tokens used per coding task vs. models with the same resultsChatty (61 ±9). Generates more tokens than models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. By effort: default 61 · high 56.Median 142,429 output tokens per task at default effort.How the Yap score worksReleased 12 Nov 2026Maker HQ: US (according to Google's own website)
All rankings: #54 on OpenRouter this week #56 on OpenRouter today #89 on OpenRouter this month #90 on NanoGPT recentlySources: Google Generative AI API data docs, LMArena data, ModelRunner.ai data, AI/ML API data, llminfo researchPrimary data from Google Generative AI API from LMArena: variant from ModelRunner.ai: weights availability, description from AI/ML API: primary mode from llminfo research: release date
Tags from its sources:instructtoolsvisionreasoningaudiodocument

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Advantages

Trade-offs

No open weights

The weights aren't public, so you can't run it yourself.

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 04:34 UTC.

Input & output

as stated by the model's providers
Text
Images
Audio
Video
Files
Text

Capabilities

Supported operations and provider-stated features

VisionReasoningFunction callingStructured outputPrompt cachingPDF inputComputer useWeb searchAudio inputBatch API

Listings

 Googlemodels/gemini-4-hydrogenSources: Google Generative AI API data docs, LiteLLM data, Models.dev data1M$1.50cache read $0.075batch $0.75$7.50batch $3.75———
 Nous Portalgoogle/gemini-4-hydrogen:batch · batchSource: Nous Portal Inference Gateway data1M$0.75cache read $0.0375$3.75$0.75—$0.00000075
 OpenRouterInference by Googlegoogle/gemini-4-hydrogen:batch · batch, globalSource: OpenRouter data docs1M$0.75cache read $0.0375$3.75$0.75$0.0375$0.00000075
 Aihubmixgemini-4-hydrogenSource: AIHubMix data docs1M$1.50cache read $0.075$7.50———
 Atlascloudgoogle/gemini-4-hydrogenSource: AtlasCloud data1M$1.50cache read $0.075$7.50———
 Cometapigemini-4-hydrogenSources: CometAPI data, CometAPI Pricing dataNot published by the provider$1.50$7.50———
 Eden AIInference by Googlegoogle/gemini-4-hydrogenSource: Models.dev data1M$1.50cache read $0.075$7.50———
 Evolinkgemini-4-hydrogenSources: EvoLink data, EvoLink PricingNot published by the provider$1.50$7.50———
 Fastroutergoogle/gemini-4-hydrogenSource: FastRouter data1M$1.50cache read $0.075batch $0.75$7.50batch $3.75———
 Kilogoogle/gemini-4-hydrogenSources: Kilo AI Gateway data, Models.dev data1M$1.50cache read $0.075$7.50——$0.00000075
Show all 38 listings

Showing 10 of 38: the maker's own listing first, then the cheapest per 1M input tokens from each provider.

Also sold by Cloudflare AI Gateway, Cortecs AI, Crossmodel, Github Copilot, Gmicloud, Llmgateway, Llmgateway Providers, Meganova, Merge Gateway, NanoGPT, Neosantara, Netmind, Ofox, Perplexity, Venice, Vercel, Vivgrid, Wavespeed, Zenmux.

9 listings without a price are hidden.

Included in plans

Copilot EnterprisegithubSubscription$39/moextra use billed past the allowancePlan page
Copilot ProgithubSubscription$10/moextra use billed past the allowancePlan page
Copilot Pro+githubSubscription$39/moextra use billed past the allowancePlan page
Copilot BusinessgithubSubscription$19/moextra use billed past the allowancePlan page
Copilot MaxgithubSubscription$100/moextra use billed past the allowancePlan page

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Human preference

Arena ELO37 related benchmarks · 37 results
Design Arena EloSource: Design Arena Leaderboards data · Nov 2026

Coding

Arena ELOSource: LMArena data · Nov 2026

Agentic

Arena AgentSource: LMArena data · Nov 2026
ValsSource: Vals AI

Reasoning

General

Math

Multilingual

Instruction following

Factuality

Science

Other

ValsSource: Vals AI

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · LMArena, CC BY 4.0 (licence) · Baromètre de l’écriture IA, Johannes Eva; public numeric ranking data, © 2026 Johannes Eva, no results licence stated · BullshitBench by Peter Gostev, MIT (licence) · Deep20Bench edition 1.1, Patrick Heusser. Results CC BY 4.0; native question scores and fraction-to-percent success rates. (licence) · MineBench, MIT (licence) · Hugging Face official benchmark leaderboards; licence varies by benchmark · UzLiB, Tahrirchi (Abror Shopulatov). Public LEADERBOARD.md numbers; repository MIT covers software and documentation, no separate results licence stated and none asserted.

Also known as: gemini 4 hydrogen, gemini-4-hydrogen-high, gemini-4-hydrogen@eu, gemini-4-hydrogen@us, gemini/gemini-4-hydrogen, github_copilot/gemini-4-hydrogen, gmi/google/gemini-4-hydrogen, google-ai-studio/gemini-4-hydrogen, google-gemini-4-hydrogen, google-vertex/gemini-4-hydrogen, google/gemini-4-hydrogen, google/gemini-4-hydrogen:batch, modelrunner-ai/google-gemini-4-hydrogen, openrouter/google/gemini-4-hydrogen, perplexity/google/gemini-4-hydrogen, vertex_ai/gemini-4-hydrogen

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works