Is Gemini 3.5 Pro available yet in July 2026?

No. As of late July 2026, Gemini 3.5 Pro is still in limited Vertex AI enterprise preview and is not generally available, having missed its June 2026 target. Google says it needs more time on coding and reliability. In the meantime Google shipped Gemini 3.6 Flash and 3.5 Flash-Lite, which are GA.

Should I use Claude Opus 5 or wait for Gemini 3.5 Pro?

Use Claude Opus 5 now. It's GA (July 24, 2026), priced at $5/$25, and leads on multi-file coding and computer use. Gemini 3.5 Pro's leaked benchmarks look strong (rumored 2M context, Deep Think layer) but it isn't broadly available, so it can't be your production default today.

What Gemini model can I use today for coding?

Gemini 3.6 Flash ($1.50/$7.50, GA July 21, 2026) and Gemini 3.5 Flash-Lite ($0.30/$2.50) are the GA options. Flash-Lite scores 54.2% on SWE-Bench Pro and 54% on Terminal-Bench 2.1 — solid for a budget tier. For frontier coding, Claude Opus 5 or GPT-5.6 Sol are the available leaders.

Quick Answer

Claude Opus 5 vs Gemini 3.5 Pro: What to Use Now

Published: July 27, 2026

The Short Answer

Use Claude Opus 5 now — Gemini 3.5 Pro still isn’t GA. Opus 5 shipped July 24, 2026 at $5/$25 with frontier-class coding. Gemini 3.5 Pro remains in limited Vertex AI preview (missed its June target). Google’s available options today are the Flash models, not Pro.

Availability (the deciding factor)

Model	Status (late July 2026)	Price (in/out)
Claude Opus 5	GA (July 24, 2026)	$5 / $25
Gemini 3.5 Pro	Preview only (Vertex enterprise)	$15 / $60 (announced)
Gemini 3.6 Flash	GA (July 21, 2026)	$1.50 / $7.50
Gemini 3.5 Flash-Lite	GA (July 21, 2026)	$0.30 / $2.50

What Each Brings

Claude Opus 5 (use it today):

1M context, $5/$25, SWE-bench Pro 79.2%, Coding Index 78.0%.
Leads multi-file/repo coding and computer use; strong long-horizon autonomy.
Broadly available: API, AWS Bedrock, Claude apps (Pro/Max/Team/Enterprise).

Gemini 3.5 Pro (not yet):

Rumored 2M-token context and a “Deep Think” inference layer for multi-step logic.
Leaked benchmarks suggest strength in zero-shot, agentic, and multi-file/tool-use tasks.
But it’s preview-gated — you can’t standardize production on it.

Gemini today = Flash tier:

Gemini 3.6 Flash — new default, up to 65% fewer output tokens than 3.5 Flash.
Gemini 3.5 Flash-Lite — 54.2% SWE-Bench Pro, 54% Terminal-Bench 2.1, at $0.30/$2.50.

What to Do

Standardize on Opus 5 (or GPT-5.6 Sol) for frontier coding/agents now.
Use Gemini 3.6 Flash as a cheap default for high-volume, simpler calls.
Pilot Gemini 3.5 Pro in Vertex preview if you have access — but don’t bet production on it until GA.
Keep model choice a config value so you can flip to Gemini 3.5 Pro the day it ships.

Sources

Anthropic — Claude Opus 5 (July 24, 2026): anthropic.com/news/claude-opus-5
CometAPI — Gemini 3.5 Pro release date & specs (updated July 2026): cometapi.com/gemini-3-5-pro-release-date-rumored-specifications-all-we-know-in-2026-updated-july-2026
getbind — Gemini 3.5 Pro slips to July: blog.getbind.co/gemini-3-5-pro-slips-to-july