Gemini 3.6 Flash
gemini-3-6-flashgemini-3.6-flashgoogle/gemini-3.6-flash- 🆕 Google's latest Flash-tier model, based on Gemini 3.5 Flash.
- 📏 One million token context, up to 64K output tokens.
- 👁️ Multimodal inputs: text, image, audio, and video.
- 🧠 Reasoning model with configurable thinking levels for cost control.
- 🔧 Built-in computer use tool for agentic execution.
- ⚡ Google reports fewer output tokens and fewer turns than predecessor.
- 🎯 Optimized for coding, agentic loops, and multi-step orchestration.
- 🏢 Available via Gemini API, Google Cloud, and Gemini app.
Google is an American multinational technology corporation and one of the world's most valuable brands. A subsidiary of parent company Alphabet Inc., Google operates across search, cloud computing, consumer electronics, and artificial intelligence. Its DeepMind and Google…
Explore 27 more models by Google →Gemini 3.6 Flash is Google's July 2026 update to the efficiency-focused Flash line, positioned between low-cost throughput models and the Pro tier. Google positions it for real-world agentic tasks—code generation, multi-step orchestration, and spatial reasoning. It carries a one-million-token context window and accepts text, image, audio, and video inputs, outputting up to 64K tokens. Google's documentation confirms it is the successor to Gemini 3.5 Flash.
Against that predecessor, Google emphasizes token efficiency: it reports 3.6 Flash uses fewer tokens and completes multi-step workflows in fewer turns, while lowering compile-failure and revision rates in coding environments. DeepMind states it executes code migrations with lower latency and higher quality than 3.5 Flash. A built-in computer use tool and configurable thinking levels (from minimal to high) let developers trade latency against reasoning depth.
It launched alongside Gemini 3.5 Flash-Lite, a higher-throughput sibling, and follows earlier Flash checkpoints including Gemini 3 Flash Preview. Independent evaluator Artificial Analysis characterizes 3.6 Flash as fast and fairly concise, and confirms its reasoning behavior and one-million-token context. For deeper reasoning, Google also offers Gemini 3.1 Pro Preview in the Pro tier.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
| Seller | Reputation↓ | Routing | Input $/M | Cached $/M | Output $/M | Categories | API |
|---|---|---|---|---|---|---|---|
| ▲ Apex Ant 0x73b4…e736 | 73.25 | #3 | $0.4688 | $0.05 | $2.34 | chat, fast, premium | openai-chat-completions |
| Edith AI 0xb269…b1a6 | 70.47 | #2 | $0.15 | $0.015 | $0.75 | chat, coding | openai-chat-completions |
| attacompute.ai 0xd5e7…7be0 | 65.45 | #4 | $1.86 | $1.86 | $9.30 | chat, reasoning, coding | openai-chat-completions |
| Hana Gateway ✅ 0x4ae1…117b | 63.30 | #1 | $0.10 | $0.10 | $0.50 | chat, coding, fast | openai-chat-completions |
| klinn 0x0ee1…6c41 | 55.13 | gated | $0.4511 | $0.0475 | $2.2555 | chat, coding | openai-chat-completions |
| Fire Ant 🔥🐜 0xbe05…bc5d | 50.00 | gated | $0.7415 | $0.0742 | $3.7077 | — | — |
| Joi Gateway 0x8a90…7007 | 27.73 | gated | $0.4118 | $0.4118 | $2.0588 | chat | openai-chat-completions |
| PawFlor-SI-Gateway 0x70bc…8261 | 25.29 | gated | $0.4725 | $0.4725 | $2.3625 | chat, fast | openai-chat-completions |
| Apex TEE Test 0xe672…7955 | 15.22 | gated | $1.3899 | $1.3899 | $6.9495 | chat, fast, premium | openai-chat-completions |
"Best price" and the seller table are live AntSeed catalog data (advertised $/1M tokens — or $ per generated image for unit-billed image models — not settled amounts). Reputation = buyer trust score (0-100, the AntSeed SDK's own formula). "Routing" = the SDK's default buyer routing (what the VPR desktop app ships with): a trust ≥ 60 gate on the effective reputation, then cheapest-first among routable sellers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on AntSeed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.