{"id":"gemini-4-hydrogen","slug":"gemini-4-hydrogen","name":"Gemini 4 Hydrogen","orgId":"google","family":"gemini-4-hydrogen","releaseDate":"2026-11-11","availability":"ga","openWeights":false,"license":"Not stated as a weight license","paramsTotal":null,"paramsSource":null,"contextWindow":1048576,"knowledgeCutoff":null,"modelCardUrl":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-hydrogen/","hfRepo":null,"modalities":["text","image","audio","video","file"],"aliases":["Gemini 4 Hydrogen","gemini-4-hydrogen","Gemini 4 Hydrogen Flash","google gemini 4 hydrogen","google deepmind gemini 4 hydrogen","gemini-4-hydrogen-flash","google/gemini-4-hydrogen","Google: Gemini 4 Hydrogen","Gemini 4 Hydrogen (High)","gemini-4-hydrogen_xhigh","gemini 4 hydrogen"],"aliasNote":null,"origin":"tracked","tracked":{"trackedSince":"2026-11-12","announced":"2026-11-12","blurb":"Google DeepMind's Gemini 4 model for everyday coding and agentic workflows, with four thinking levels (low, medium, high by default, and xhigh). Announced November 12, 2026. Available through the Gemini API and Google AI Studio as gemini-4-hydrogen, and rolling out in the Gemini app, Google Search, Gemini for business, and paid Google AI Pro and Ultra paths in Google Antigravity. Priced at $1.50 per 1M input tokens and $7.50 per 1M output tokens, flat across a 1 million-token context window. No public weight release is stated.","api":"live","apiModelId":"gemini-4-hydrogen","openWeights":"no","facts":[{"label":"Type","value":"Gemini 4 model for everyday coding and agentic workflows, including as an execution model alongside Gemini 4 Carbon and Gemini 4 Argon in multi-model setups such as Orchestrator in Google Antigravity (Google blog)"},{"label":"Thinking levels","value":"Four official levels: low, medium, high (default), and xhigh (Google blog)"},{"label":"Context","value":"Official docs: unified context window of 1,048,576 tokens, with a default synchronous max output of 65,536 tokens and long decode up to 1,048,576 tokens. The blog also describes a 1 million-token input window and up to 1 million output tokens (Google docs and blog)"},{"label":"Modalities","value":"Inputs: text, image, video, audio, and PDF. Output: text. Official docs list thinking, function calling, code execution, computer use, file search, structured outputs, URL context, implicit context caching, and grounding with Google Search and Maps. Image generation, audio generation, and the Live API are not supported (Google docs)"},{"label":"Hosted access","value":"Live in the Gemini API and Google AI Studio, model id gemini-4-hydrogen. Official blog: also rolling out in the Gemini app, Google Search, Gemini for business, Gemini in Google Sheets for eligible subscriptions, paid-tier Android Studio, and Google Antigravity for paid Google AI Pro and Ultra subscribers. In the Gemini app, free users get it via automatic model selection when it is the appropriate model for the prompt (Google blog and docs)"},{"label":"Pricing","value":"Paid standard tier: $1.50 per 1M input tokens and $7.50 per 1M output tokens, including thinking tokens, flat across the 1 million-token context window. Free-tier input and output are free of charge, with free-tier prompts used to improve Google products. Implicit context caching is $0.075 per 1M tokens on the paid tier (explicit caching is not supported). Batch and Flex paid tiers are $0.75 / $3.75; Priority is $2.70 / $13.50 (Google blog and Gemini Developer API pricing)"},{"label":"Weights","value":"No public weight release stated on the official blog or docs"},{"label":"License","value":"Not stated as a weight license. Access is under the Gemini API / Google AI Studio terms"},{"label":"Related","value":"Part of the Gemini 4 family with Gemini 4 Carbon and Gemini 4 Argon; listed on the Gemini 4 family hub. Separate from Deep Principle's Hydrogen-1.0"}],"claims":["Lab claim: self-reported Artificial Analysis Intelligence Index score of 51 at the default high thinking level, versus 50 for Gemini 4 Argon, and 54 at the xhigh thinking setting (Google blog).","Lab claim: self-reported as finishing tasks 2.5 times faster than Gemini 3.8 Flash at its high default setting, with Terminal-Bench 4.1 scores of 57% versus 18.2% (Google blog).","Lab claim: self-reported as cutting fabricated claims by more than 5 times versus Gemini 3.8 Flash, and as reducing overreach in agentic tasks by up to 81% and under-service by up to 79% (Google blog)."],"pitch":"a Gemini 4 model for everyday coding and agentic workflows, available in the Gemini API, Google AI Studio, the Gemini app and Search","sources":[{"kind":"lab","title":"Introducing Gemini 4 Hydrogen (Google blog)","url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-hydrogen/","date":"2026-11-12"},{"kind":"lab","title":"Gemini 4 Hydrogen model card (Gemini API docs)","url":"https://aistudio.google.com/docs/models/gemini-4-hydrogen"},{"kind":"lab","title":"Gemini Developer API pricing","url":"https://aistudio.google.com/docs/pricing"},{"kind":"lab","title":"Google DeepMind on X: Gemini 4 Hydrogen","url":"https://x.com/GoogleDeepMind/status/2120940338869055788","date":"2026-11-12"},{"kind":"lab","title":"Sundar Pichai on X: Gemini 4 Hydrogen","url":"https://x.com/sundarpichai/status/2120940236805169182","date":"2026-11-12"},{"kind":"lab","title":"Gemini app on X: Gemini 4 Hydrogen","url":"https://x.com/GeminiApp/status/2120940601319260164","date":"2026-11-12"}],"firstBenchmark":{"date":"2026-11-13","benchmarkId":"vals-index"}},"contextSource":"openrouter","listedSince":"2026-11-11","eligibleSince":"2026-11-13","organization":{"id":"google","name":"Google DeepMind","slug":"google","country":null,"url":"https://deepmind.google","color":"#4285F4","mark":"G"},"price":{"modelId":"gemini-4-hydrogen","provider":"google","inputPerM":null,"outputPerM":null,"blendedPerM":3,"tokensPerS":null,"observedAt":"2026-11-22","sourceId":"openrouter"},"answer":{"word":"Yes.","scope":"Yes on Factuality (best per dollar). Not the overall leader.","lead":"As of Nov 22, 2026, Gemini 4 Hydrogen is state of the art on best per dollar in Factuality (since Nov 13, 2026; sources LMArena Text factuality; methodology v1.0). Gemini 4 Hydrogen shares the Factuality best-per-dollar frontier with DeepSeek V4.1 Flash, Gemini 4 Carbon, GLM 5.3, Granite 4.2 8B, Mimo V2.6 Pro since Nov 13, 2026. Co-SOTA shares rank 1.","confidence":"medium"},"status":"BENCHMARKED","verdicts":[{"id":"v_overall_general-llm_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"general-llm","category_name":"General LLM","slice":"overall","state":"RANKED","rank":14,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":8.771929824561397,"confidence":"medium","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"Ranked #14 in General LLM.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#14 in General LLM"},{"id":"v_pareto_general-llm_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"general-llm","category_name":"General LLM","slice":"pareto","state":"RANKED","rank":14,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":8.771929824561397,"confidence":"medium","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"Ranked #14 in General LLM on best per dollar.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#14 in General LLM (best per dollar)"},{"id":"v_overall_reasoning_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"reasoning","category_name":"Reasoning","slice":"overall","state":"RANKED","rank":17,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":10.588235294117652,"confidence":"medium","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"Ranked #17 in Reasoning.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#17 in Reasoning"},{"id":"v_pareto_reasoning_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"reasoning","category_name":"Reasoning","slice":"pareto","state":"RANKED","rank":17,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":10.588235294117652,"confidence":"medium","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"Ranked #17 in Reasoning on best per dollar.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#17 in Reasoning (best per dollar)"},{"id":"v_overall_science_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"science","category_name":"Science","slice":"overall","state":"RANKED","rank":2,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":null,"confidence":"low","valid_from":"2026-11-22","valid_to":null,"configuration":"xhigh","rationale":"Last listed at #2 in Science on 2026-11-12. That board does not award a crown.","safeguard_flags":["off_board"],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#2 in Science"},{"id":"v_overall_coding_gemini-4-hydrogen_2026-11-22_open_INSUFFICIENT_DATA","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"coding","category_name":"Coding","slice":"overall","state":"INSUFFICIENT_DATA","rank":null,"present_evidence":[{"benchmark":"lmarena-text-coding","name":"LMArena Text (style-controlled, coding)","rank":50}],"missing_evidence":[{"benchmark":"terminal-bench-4","name":"Terminal-Bench 4.0"},{"benchmark":"swe-bench-pro-hard","name":"SWE-bench Pro V2 HARD"}],"coverage":null,"segments":[],"confirm":[],"margin":null,"confidence":"low","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"#50 on LMArena Text (style-controlled, coding); missing Terminal-Bench 4.0 and SWE-bench Pro V2 HARD","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"Not enough data"},{"id":"v_overall_coding-web-dev_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"coding-web-dev","category_name":"Coding › Web dev","slice":"overall","state":"RANKED","rank":12,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":8.759124087591246,"confidence":"medium","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"Ranked #12 in Coding › Web dev.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#12 in Coding › Web dev"},{"id":"v_pareto_coding-web-dev_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"coding-web-dev","category_name":"Coding › Web dev","slice":"pareto","state":"RANKED","rank":12,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":8.759124087591246,"confidence":"medium","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"Ranked #12 in Coding › Web dev on best per dollar.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#12 in Coding › Web dev (best per dollar)"},{"id":"v_overall_long-context_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"long-context","category_name":"Long context","slice":"overall","state":"RANKED","rank":9,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":5.88235294117645,"confidence":"medium","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"Ranked #9 in Long context.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#9 in Long context"},{"id":"v_pareto_long-context_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"long-context","category_name":"Long context","slice":"pareto","state":"RANKED","rank":9,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":5.88235294117645,"confidence":"medium","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"Ranked #9 in Long context on best per dollar.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#9 in Long context (best per dollar)"},{"id":"v_overall_vision_gemini-4-hydrogen_2026-11-22_open_INSUFFICIENT_DATA","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"vision","category_name":"Vision","slice":"overall","state":"INSUFFICIENT_DATA","rank":null,"present_evidence":[{"benchmark":"gdp-pdf","name":"GDP.pdf","rank":3}],"missing_evidence":[{"benchmark":"lmarena-vision","name":"LMArena Vision (style-controlled)"}],"coverage":null,"segments":[],"confirm":[],"margin":null,"confidence":"low","valid_from":"2026-11-22","valid_to":null,"configuration":"xhigh","rationale":"#3 on GDP.pdf; missing LMArena Vision (style-controlled)","safeguard_flags":["off_board"],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"Not enough data"},{"id":"v_overall_factuality_gemini-4-hydrogen_2026-11-22_open_NEAR_SOTA","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"factuality","category_name":"Factuality","slice":"overall","state":"NEAR_SOTA","rank":5,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":2.941176470588232,"confidence":"medium","valid_from":"2026-11-22","valid_to":null,"configuration":null,"rationale":"Close in Factuality: rank #5, 2.9 percentile points behind the leader.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"Near-SOTA · Factuality"},{"id":"v_pareto_factuality_gemini-4-hydrogen_2026-11-13_open_CO_SOTA","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"factuality","category_name":"Factuality","slice":"pareto","state":"CO_SOTA","rank":1,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":0,"confidence":"medium","valid_from":"2026-11-13","valid_to":null,"since_text":"since Nov 13, 2026","configuration":null,"rationale":"Gemini 4 Hydrogen shares the Factuality best-per-dollar frontier with DeepSeek V4.1 Flash, Gemini 4 Carbon, GLM 5.3, Granite 4.2 8B, Mimo V2.6 Pro since Nov 13, 2026. Co-SOTA shares rank 1.","safeguard_flags":["pareto","two_signal_confirmed"],"co_sota_with":["deepseek-v4-1-flash","gemini-4-carbon","glm-5-3","granite-4-2-8b","mimo-v2-6-pro"],"challengers":[],"methodology":"1.0","label":"Pareto · Factuality"},{"id":"v_overall_legal_gemini-4-hydrogen_2026-11-22_open_RANKED","model":"gemini-4-hydrogen","model_name":"Gemini 4 Hydrogen","category":"legal","category_name":"Legal","slice":"overall","state":"RANKED","rank":4,"present_evidence":[],"missing_evidence":[],"coverage":null,"segments":[],"confirm":[],"margin":null,"confidence":"low","valid_from":"2026-11-22","valid_to":null,"configuration":"xhigh","rationale":"Ranked #4 in Legal. That board does not award a crown.","safeguard_flags":[],"co_sota_with":[],"challengers":[],"methodology":"1.0","label":"#4 in Legal"}]}