{"slug": "gemini-3-6-flash-3-5-flash-lite-developer-guide", "title": "Gemini 3.6 Flash & 3.5 Flash-Lite: Developer guide", "summary": "Google has made Gemini 3.6 Flash and Gemini 3.5 Flash-Lite generally available for production use. Gemini 3.6 Flash offers stronger performance on complex agentic and multimodal tasks with reduced token usage at a lower price than 3.5 Flash, while Gemini 3.5 Flash-Lite is the fastest, lowest-cost model in the 3.5 family for high-throughput execution.", "body_md": "Gemini 3.6 Flash (`gemini-3.6-flash`\n\n) and Gemini 3.5 Flash-Lite (`gemini-3.5-flash-lite`\n\n) are generally available (GA) and ready for production use.\n\n-\n**Gemini 3.6 Flash**: Stronger performance on complex agentic and multimodal tasks while reducing token usage, at a lower price point than 3.5 Flash.\n-\n**Gemini 3.5 Flash-Lite**: The fastest, lowest-cost model in the 3.5 family. Outperforms prior Flash-Lite generations for high-throughput execution.\n\nThis guide explains what's new in each model, what API changes affect your code, and how to migrate.\n\n**Tip**: You can automate this migration with a coding agent that supports [skills](https://ai.google.dev/gemini-api/docs/coding-agents#gemini-interactions-api) (like Antigravity). Run `/gemini-interactions-api migrate my app to Gemini 3.6 Flash`\n\n##\nNew models\n\n| Model |\nModel ID |\nDefault thinking level |\nPricing |\nDescription |\n| Gemini 3.6 Flash |\n`gemini-3.6-flash` |\n`medium` |\n$1.50/1M input tokens and $7.50/1M output tokens |\nBalances speed with intelligence for agentic and multimodal tasks. |\n| Gemini 3.5 Flash-Lite |\n`gemini-3.5-flash-lite` |\n`minimal` |\n$0.30/1M input tokens and $2.50/1M output tokens |\nThe fastest, lowest-cost 3.5 model for high-throughput execution. |\n\nBoth models support the 1M token context window, 64k max output tokens, thinking, and the full suite of built-in tools including [Computer Use](https://ai.google.dev/gemini-api/docs/computer-use).\n\nFor complete specs, see the model pages:\n\nFor detailed pricing, see the [pricing page](https://ai.google.dev/gemini-api/docs/pricing).\n\n##\nQuickstart\n\n##\nWhat's new in Gemini 3.6 Flash\n\n-\n**Token and turn reduction:** Completes multi-step workflows with fewer reasoning steps, conversational turns, and tool calls than Gemini 3.5. It also reduces execution loop spiraling.\n-\n**Improved code generation:** Produces higher quality production-ready code with fewer unwanted edits and fewer debugging loops.\n-\n**Better instruction following**: Reduces unwanted file changes during diagnostic tasks.\n-\n**Strong multimodal and spatial reasoning:** Improved performance on chart interpretation, visual blueprint conversion, and multi-element web layout generation.\n-\n**Upfront programmatic inspection:** Prefers running diagnostic code scripts before making changes more frequently than Gemini 3.5 Flash. This improves accuracy on complex tasks, but can add extra exploratory steps on simple frontend work.\n-\n**Computer Use support:** Supported as native tool for agentic UI automation.\n-\n**UI styling preference**: Better at creating functional code, though human evaluators preferred earlier models for visual layout and styling. You can mitigate this by providing explicit design guidelines.\n-\n**Default thinking effort (medium):** Uses the same `medium`\n\ndefault thinking level as Gemini 3.5 Flash.\n-\n**Reduced pricing**: Lower output token costs ($7.50/1M vs. $9.00/1M for 3.5 Flash). Input tokens remain at $1.50/1M.\n\n##\nWhat's new in Gemini 3.5 Flash-Lite\n\n-\n**Reduced task execution latency:** Highest throughput in the 3.5 family for high-volume data parsing and document extraction.\n-\n**Enhanced reasoning and multimodal performance:** Strong migration path from Gemini 2.5 Flash, with higher scores on reasoning tasks like HLE (18.0% vs. 11.0%) and multimodal benchmarks like CharXIV (74.5% vs. 63.7%).\n-\n**Subagent orchestration and tool reliability:** Improves tool execution reliability for code execution, search, and MCP workflows. Increase the thinking level for autonomous planning and complex subagent tasks.\n-\n**Improved document understanding:** Improves accuracy on document parsing and structured data extraction. Experiment with both minimal and high thinking levels depending on document complexity.\n-\n**Interactive web coding and tabular data processing:** Performs strongly on frontend JavaScript and tabular data processing by planning via lightweight code execution.\n-\n**Chatbot and persona persistence:** Stronger multi-turn instruction following and persona consistency over Gemini 3.1 Flash-Lite.\n-\n**Computer Use support:** Supported as native tool for agentic UI automation.\n\n##\nChoosing the right Flash or Flash-Lite model\n\nUse this table to select the right model and migration path for your workloads.\n\nBoth models require removing deprecated sampling parameters (`temperature`\n\n, `top_p`\n\n, `top_k`\n\n) and prefilled model turns. See API changes for details.\n\n| Model |\nPrimary use cases |\nRecommended migration target |\n**Gemini 3.6 Flash**\n`gemini-3.6-flash`\n|\nCode generation, spatial/multimodal reasoning, multi-step agentic workflows |\n**Gemini 3.5 Flash**, **Gemini 3 Flash (Preview)**, or **Gemini 3.1 Pro**\n|\n**Gemini 3.5 Flash-Lite**\n`gemini-3.5-flash-lite`\n|\nAutonomous subagent execution, high-volume data analysis and document extraction, structured JSON parsing |\n**Gemini 3.1 Flash-Lite** or **Gemini 2.5 Flash**\n|\n\n##\nUpdated Antigravity agent\n\nDue to its improved performance, Gemini 3.6 Flash is now the new default model powering the [Antigravity agent](https://ai.google.dev/gemini-api/docs/antigravity-agent) in Gemini Managed Agents. This can be changed by setting a new field on the API.\n\n##\nAPI changes and parameter updates\n\nStarting with Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, the following API changes apply to these models and all future Gemini model releases.\n\n-\n**Sampling parameter deprecation**: `temperature`\n\n, `top_p`\n\n, and `top_k`\n\nare deprecated. The API ignores these parameters and returns an error in future model generations.\n-\n**Prefilled model turn validation**: Prefilling model turns is no longer supported. If the last non-empty turn in the request is a `model`\n\nturn, the API returns a `400`\n\nerror.\n\nBelow are detailed explanations and code samples for each API change.\n\n###\n1. Sampling parameter deprecation (`temperature`\n\n, `top_p`\n\n, `top_k`\n\n)\n\n`temperature`\n\n, `top_p`\n\n, and `top_k`\n\nare deprecated and ignored. In future model generations, supplying these parameters returns an HTTP 400 error. **Remove these parameters from all requests.**\n\nTo improve determinism, define a system instruction with explicit rules for your\n\nspecific use case.\n\n###\n2. Prefilled model turn validation {:#prefilled-model-turn-validation}\n\nAPI requests ending with a non-empty model role turn are disallowed and return an **HTTP 400 Error**.\n\n####\n⚠️ Avoid\n\nIn legacy `generateContent`\n\nor raw REST payloads, ending with a model role turn\n\nis now disallowed:\n\n####\n✅ Recommended Migration (Interactions API)\n\nIn the Interactions API, model turns are not manually prefilled. If your application previously prefilled a model turn to suppress preambles or force JSON formatting, use system_instruction or [Structured outputs](https://ai.google.dev/gemini-api/docs/structured-output) instead.\n\n##\nMigration checklist\n\n**Tip**: You can automate this migration with a coding agent that supports [skills](https://ai.google.dev/gemini-api/docs/coding-agents#gemini-interactions-api) (like Antigravity). Run `/gemini-interactions-api migrate my app to Gemini 3.6 Flash`\n\n###\nMigrate to gemini-3.6-flash\n\n-\n**Update Model ID:** Change your target model string to `gemini-3.6-flash`\n\n.\n-\n**Remove deprecated sampling parameters:**\n- Strip\n`temperature`\n\n, `top_p`\n\n, and `top_k`\n\nfrom generation configs.\n- Replace\n`thinking_budget`\n\nwith the string enum `thinking_level`\n\nset to `\"medium\"`\n\nor `\"high\"`\n\n.\n- Remove\n`candidate_count`\n\n(unsupported in Gemini 3.x).\n\n-\n**Enforce turn validation rules:**\n- Standardize multi-turn conversations on server-side\n`previous_interaction_id`\n\n.\n- Remove prefilled model turns.\n\n-\n**Audit function calling:**\n- Place multimodal assets inside the response payload.\n- Format inline instructions using\n`\\n\\n`\n\n.\n- If you see\n`Malformed_Function_Call`\n\nerrors tied to pre-tool text, see [Workarounds for pre-tool text requirements](https://ai.google.dev/gemini-api/docs/function-calling#workarounds-for-pre-tool-text-requirements).\n- Only if using generateContent API: Ensure all\n`FunctionResponse`\n\nobjects include `call_id`\n\nand `name`\n\n.\n\n-\n**Baseline Gemini 3.x requirements:** For SDK updates and thought signature preservation, see the [Gemini 3.5 Migration Checklist](https://ai.google.dev/gemini-api/docs/whats-new-gemini-3.5#migration).\n\n###\nMigrate to gemini-3.5-flash-lite\n\n-\n**Update Model ID:** Change your target model string to `gemini-3.5-flash-lite`\n\n.\n-\n**Configure thinking effort level:**\n- For high-volume extraction, routing, or classification: leave\n`thinking_level`\n\nat `\"minimal\"`\n\n(default) for maximum throughput.\n- For autonomous subagents with tool calls, code execution, or multi-step reasoning: set\n`thinking_level`\n\nto `\"medium\"`\n\nor `\"high\"`\n\nto prevent premature tool termination.\n\n-\n**Remove deprecated parameters and validate function calling:** Apply the same rules as for 3.6 Flash above.\n-\n**Baseline Gemini 3.x requirements:** Refer to the [Gemini 3.5 Migration Checklist](https://ai.google.dev/gemini-api/docs/whats-new-gemini-3.5#migration).\n\n##\nNext steps", "url": "https://wpnews.pro/news/gemini-3-6-flash-3-5-flash-lite-developer-guide", "canonical_source": "https://dev.to/googleai/gemini-36-flash-35-flash-lite-developer-guide-i17", "published_at": "2026-07-21 16:08:54+00:00", "updated_at": "2026-07-21 16:23:05.205100+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-products", "ai-infrastructure"], "entities": ["Google", "Gemini 3.6 Flash", "Gemini 3.5 Flash-Lite"], "alternates": {"html": "https://wpnews.pro/news/gemini-3-6-flash-3-5-flash-lite-developer-guide", "markdown": "https://wpnews.pro/news/gemini-3-6-flash-3-5-flash-lite-developer-guide.md", "text": "https://wpnews.pro/news/gemini-3-6-flash-3-5-flash-lite-developer-guide.txt", "jsonld": "https://wpnews.pro/news/gemini-3-6-flash-3-5-flash-lite-developer-guide.jsonld"}}