Is Gemini 3.5 Pro out yet in August 2026?

No. As of August 5, 2026, Gemini 3.5 Pro is still NOT generally available — Google's Pro page shows 'coming soon' and it remains in limited Vertex preview. Leaks point to an August 12, 2026 launch. Announced pricing is $15/$60 per MTok. Meanwhile Alibaba's Qwen3.8-Max shipped on August 3 with open weights promised the following week.

Should I wait for Gemini 3.5 Pro or use Qwen3.8-Max?

If you need something today, don't wait on Gemini 3.5 Pro — use a proven model (Claude Opus 5, GPT-5.6 Sol) or try Qwen3.8-Max once its open weights drop (~week of Aug 10, 2026). Gemini 3.5 Pro's $15/$60 pricing is also 3x Opus 5's output cost, so weigh that against its (still unproven at GA) capabilities.

How do Qwen3.8-Max and Gemini 3.5 Pro compare on price?

Gemini 3.5 Pro's announced API price is $15 input / $60 output per MTok — the priciest frontier tier. Qwen3.8-Max API pricing was TBD at its August 3 launch, but its open weights let you self-host instead. For cheap frontier work, Gemini 3.6 Flash ($1.50/$7.50) is the better Google option today.

Quick Answer

Qwen3.8-Max vs Gemini 3.5 Pro (Aug 2026)

Published: August 5, 2026

The Short Answer

Qwen3.8-Max actually shipped (Aug 3, 2026, open weights ~week of Aug 10). Gemini 3.5 Pro is still not GA as of August 5 — leaked for August 12 at a steep $15/$60. If you need a frontier model now, Qwen3.8-Max (or a proven closed model) beats waiting on Google.

Side-by-Side

	Qwen3.8-Max	Gemini 3.5 Pro
Vendor	Alibaba	Google
Status (Aug 5, 2026)	Shipped Aug 3; open weights ~Aug 10	NOT GA — limited preview, leaked Aug 12
Type	Open-weight	Closed
Params	2.4T (MoE)	Undisclosed
API price (in/out per MTok)	TBD	$15 / $60 (announced)
GPQA Diamond*	92.6	— (not GA)

*Alibaba-reported, unverified independently as of Aug 5, 2026.

How To Choose

Need a frontier model this week → Qwen3.8-Max (once weights ship) or a proven closed model. Gemini 3.5 Pro keeps slipping — it was announced at Google I/O May 19, moved to June, then July, and is still preview-only.
Cheap Google frontier today → skip 3.5 Pro’s $15/$60 and use Gemini 3.6 Flash ($1.50/$7.50), Google’s current default.
Self-hosting / data control → Qwen3.8-Max is the only option here — Gemini is API-only.

The Trade-Off

Shipped vs vaporware-adjacent: Qwen3.8-Max is real code you can (soon) download. Gemini 3.5 Pro is still a preview after three delays.
Price: Gemini 3.5 Pro’s $15/$60 is 3x Opus 5’s output. Qwen3.8-Max lets you dodge per-token cost via self-hosting.
Proof: Neither has fully independent public benchmarks at GA — but at least Qwen3.8-Max’s weights will let the community test them.

Watch Outs

Gemini 3.5 Pro dates are leaks — August 12 is unconfirmed by Google.
Qwen3.8-Max benchmarks are vendor claims and its open weights hadn’t shipped at Aug 3 launch (verify the drop + license).

Verdict

Available now / open → Qwen3.8-Max
Wait only if you specifically need Google → Gemini 3.5 Pro (leaked Aug 12), else use Gemini 3.6 Flash today

Sources

SCMP — Qwen3.8-Max availability: scmp.com
Business Insider — Gemini 3.5 Pro delays: businessinsider.com
Google — Gemini models: deepmind.google/technologies/gemini