Google ships 3 budget Gemini models, flagship still missing

Key Takeaways

• Google launched three cost-focused Gemini models: 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber for cybersecurity
• Gemini 3.5 Pro remains delayed past its June target, reportedly due to underperformance in coding tasks
• Rivals Anthropic and OpenAI have pulled ahead with Mythos 5, Fable 5, and GPT-5.6 releases

Google released three new Gemini AI models on Tuesday, expanding its lineup of cheaper, lightweight options. The flagship Gemini 3.5 Pro, however, remains conspicuously absent. The company offered no timeline beyond "soon," leaving enterprise customers and Wall Street waiting.

The new releases include Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, both updated versions of Google's budget-tier offerings. A third model, Gemini 3.5 Flash Cyber, targets cybersecurity applications specifically. Google DeepMind described it as "a cost-efficient and highly capable alternative to large, costly cybersecurity models."

Why is Gemini 3.5 Pro delayed?

CEO Sundar Pichai promised a June launch for Gemini 3.5 Pro during Google I/O in May. That deadline has slipped by several weeks. The reason, according to reports: the model is falling short of internal benchmarks, particularly in coding tasks.

Coding has become the most lucrative enterprise AI use case. Companies are paying premium prices for models that can generate, debug, and explain code reliably. If Gemini 3.5 Pro can't match or beat competitors on this front, Google risks losing ground in a market where it recently reclaimed momentum.

Google said the model is currently in partner testing and declined to provide a specific release date. Pichai will likely face questions about the delay when Alphabet reports second-quarter earnings on Wednesday.

The competitive pressure mounting on Google

Google's Gemini 3 launch last November was a genuine win. It put Google back at the front of the AI race after years of playing catch-up. OpenAI reportedly issued an internal "code red" to accelerate work on GPT in response.

That lead hasn't lasted. Anthropic shipped its Mythos 5 and Fable 5 models. The U.S. government deemed them powerful enough to pose a national security risk. OpenAI released GPT-5.6 earlier this month after its own government-mandated security review.

Google, meanwhile, is shipping incremental Flash updates while its top-tier model sits in testing. The longer Gemini 3.5 Pro stays unreleased, the more enterprise contracts go to Anthropic and OpenAI.

Google's counterargument: price

Pichai has leaned heavily on cost savings as a differentiator. He claims companies could save upward of $1 billion per year by shifting AI workloads to Gemini models. The Flash and Flash-Lite variants fit this narrative, offering cheaper inference for applications that don't require cutting-edge capabilities.

That's a real value proposition for certain workloads. Customer service bots, document summarization, and basic code completion don't always need the most powerful model. But enterprises making strategic AI investments often want the best available option, not just the cheapest.

Google also announced that training has begun on Gemini 4. That's a signal to investors that the company is thinking beyond the current delay. Whether it reassures anyone depends on how long 3.5 Pro takes to ship.

What the new models actually offer

Gemini 3.6 Flash updates the previous Flash model with unspecified performance improvements. Google hasn't published detailed benchmarks yet. Gemini 3.5 Flash-Lite sits below Flash in the lineup, presumably trading some capability for even lower cost.

The Gemini 3.5 Flash Cyber variant is the most interesting addition. Cybersecurity teams typically rely on expensive, specialized models for threat detection and analysis. A cost-efficient alternative could open AI-assisted security to smaller organizations that couldn't justify the expense before.

Frequently Asked Questions

When will Gemini 3.5 Pro be released?

Google has not provided a specific date. The company says the model is in partner testing and will arrive "soon." The original June target has already passed.

Why is Gemini 3.5 Pro delayed?

Reports indicate the model is underperforming on internal benchmarks, particularly for coding tasks. Google has not officially confirmed the specific reasons.

What are the new Gemini models Google released?

Google launched Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber. All three are lightweight, cost-focused models for different use cases.

How does Gemini compare to GPT-5.6 and Claude?

Without Gemini 3.5 Pro available, Google's current offerings lag behind OpenAI's GPT-5.6 and Anthropic's Mythos 5 in terms of top-tier capabilities.

Source: Tech-Economic Times / ET

Google reveals Flash model specs; 3.5 Pro coming soon

The new article provides detailed specifications and performance metrics for the three models: 3.6 Flash reduces output token usage by 17% (up to 65% on some benchmarks), 3.5 Flash-Lite delivers 350 output tokens per second, and 3.5 Flash Cyber is paired with CodeMender for cybersecurity. It also confirms Gemini 3.5 Pro is testing with partners and will be broadly available soon.