Gemini 3.6 Flash

Gemini 3.6 Flash launched on July 21, 2026 as a stable, production-ready model in Google's Gemini API, with the stable model ID gemini-3.6-flash. Google built it on Gemini 3.5 Flash and positioned it as a workhorse model for developers handling code, documents, images, audio, video, and tool-driven workflows, reporting The model accepts text, image, video, audio, and PDF as inputs and outputs text, supporting up to 1,048,576 input tokens and 65,536 output tokens. Google reported that it focuses on coding, knowledge work, visual tasks, and token efficiency, with distribution also covering enterprise and agent development channels. As

A third-party aggregation dated August 13, 2026, reviews documented Gemini Flash releases through the 3.6 generation and explicitly names Gemini 3.6 Flash as part of Google's July 2026 release wave alongside Gemini 3.5 Flash-Lite and Gemini 3.5 Flash Cyber. The piece tracks the Flash family rollout across the Gemini AP The aggregation is useful for governance and procurement framing: it argues organizations evaluating model versions should rely only on documented model cards, release notes, and access details from the vendor, and characterizes Gemini 3.6 Flash as a verified, production-referenceable version on the supplied record, wh

A head-to-head comparison piece dated July 29, 2026, that explicitly names Gemini 3.6 Flash and reports its benchmark scores cross-referenced from OpenAI's GPT-5.6 launch page and Google DeepMind's Gemini 3.6 Flash product page. Coding and agentic benchmarks for Gemini 3.6 Flash include SWE-Bench Pro at 58.7%, DeepSWE The piece further reports knowledge and reasoning scores for Gemini 3.6 Flash including GDPval-AA v2 Elo of 1,421 and AA Intelligence Index of 50.0, alongside GPT-5.6 Terra's higher 1,593 Elo and 55.0 Index. Pricing is presented for production-scale comparison between the two mid-tier models. The numbers are sourced to

Memeburn's benchmarks and pricing guide reports that Gemini 3.6 Flash and Gemini 3.5 Flash both score 50 on the Artificial Analysis Intelligence Index, meaning the new release matches its predecessor in measured intelligence rather than surpassing it. Average task time dropped from 2.7 minutes to 1.3 minutes, while est The piece places Gemini 3.6 Flash in a competitive context: GPT-5.6 Luna remains cheaper at public API rates, while Claude Sonnet 5 costs roughly twice as much per output token. The clearest migration case identified is for existing Gemini 3.5 Flash users whose workloads are sensitive to latency and output volume. Goog

This third-party recap confirms Gemini 3.6 Flash reached general availability in the Gemini API on July 21, 2026 at $1.50 per million input tokens and $7.50 per million output tokens, quoting Google's pricing announcement. It frames the release as Google's response in the price-sensitive, high-volume mid-tier segment a The article documents the wide simultaneous rollout: 3.6 Flash and 3.5 Flash-Lite are live in the Gemini API through Google AI Studio and Android Studio, inside Google Antigravity for developers, across the Gemini Enterprise Agent Platform and Enterprise app, and in the Gemini app where 3.6 Flash is described as availa

Google released Gemini 3.6 Flash on July 21, 2026 as the new default workhorse model in the Gemini family and the successor to Gemini 3.5 Flash, alongside simultaneous launches of Gemini 3.5 Flash-Lite and the access-restricted Gemini 3.5 Flash Cyber. It keeps the 1 million-token context window from its predecessor, ad Gemini 3.6 Flash runs at approximately 280 tokens per second, placing it among the faster models in its class for interactive use. The article notes that Google did not release Gemini 3.5 Pro in this announcement because it fell short of internal expectations on coding and complex reasoning, and that Gemini 4 was tease

Coursiv's launch summary states that Google made Gemini 3.6 Flash and Gemini 3.5 Flash-Lite generally available as production models on July 21, 2026, alongside Gemini 3.5 Flash Cyber for finding and fixing software vulnerabilities. It explicitly names "Gemini 3.6 Flash" and reports 17% fewer output tokens than 3.5 Fla The piece also details Gemini 3.5 Flash-Lite at 350 output tokens per second, $0.30/$2.50 per million tokens in/out, and the same 1M-token context and 64k output cap, noting a Search rollout and Android Studio access, while Gemini 3.5 Flash Cyber is described as a limited pilot for governments and trusted partners with

VentureBeat reports that Google DeepMind released Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber on July 21, 2026, framing the trio as token-efficient, lower-latency models aimed at scaling AI agents. The article explicitly names "Gemini 3.6 Flash" and states Google's API pricing at $1.50 per milli The piece provides independent corroboration of Gemini 3.6 Flash's positioning as a workhorse agentic and coding model with reduced output-token costs, while noting Google's broader pricing strategy relative to peers. Vendor pricing and API availability described are Google's, not Requesty's, and creative attribution i

MarkTechPost's recap confirms Google's release of Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber on July 21, 2026, all sitting in the Flash tier tuned for speed, cost, and high-volume agentic work. It explicitly names "Gemini 3.6 Flash" as the new default workhorse, reporting 17% fewer output token The article frames the lineup as targeting developers building production agents who prioritize efficiency over maximum reasoning depth, covering coding, knowledge work, and multimodal tasks. Attribution remains with Google, and the pricing figures cited are those of the official Google Gemini API rather than any Reque

TechCrunch reports that on July 21, 2026, Google DeepMind released Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber. Gemini 3.6 Flash is described as Google's workhorse model, promising improved capabilities in coding, knowledge work, and multimodal performance while reducing token usage by up to 17% The launch is contextualized by what Google did not ship: the long-anticipated update to the Gemini Pro flagship. The article notes that since the last Gemini Pro update in February, OpenAI has released GPT-5.5 and begun rolling out GPT-5.6, while Anthropic has launched Claude Opus 4.8, Claude Sonnet 5, and expanded ac

9to5Google's July 21, 2026 report independently corroborates the Gemini 3.6 Flash launch and adds concrete benchmark and pricing detail. It states 3.6 Flash consumes 17% fewer output tokens than 3.5 Flash on the Artificial Analysis Index while taking fewer reasoning steps and tool calls, and is priced at $1.50 per mill The same article also covers the sibling Gemini 3.5 Flash-Lite, targeting high-throughput and low-latency tasks like agentic search and document processing, priced at $0.30 per million input tokens and $2.50 per million output tokens, with Terminal-Bench 2.1 scores of 54% vs 31% versus the prior Flash-Lite. It further

LLM Stats' model page for Gemini 3.6 Flash provides independent third-party benchmark aggregation, explicitly naming the model "Gemini 3.6 Flash" with an overall rank of 40 on the LLM Stats composite score of 43.3 and a reported blended price of $1.79 per million tokens. Per-benchmark measurements cited include a CharX The page also lists a broader set of benchmark results drawing from each model's scorecard, paper, or official blog posts, and situates Gemini 3.6 Flash against peers such as Gemma 4 E4B, GPT OSS 120B, DeepSeek-V4-Flash-0731, and GPT-6 Astra on a log-scale cost-efficiency chart. All evaluations referenced are attribute