← All models

Gemini 3.5 Flash-Lite

Language Available Full comparison ↗

google-deepmind/gemini-3.5-flash-lite · by Google DeepMind · decoder-only-transformer

Pricing — 1 offering(s)

Google DeepMind

About this offering: Google's Gemini 3.5 Flash-Lite at $0.30 input and $2.50 output per 1M tokens, the same input rate for text, image, video and audio. It's Google's low-latency, low-cost model for high-volume agentic and data-processing work. From 2026-12-01, Google routes Gemini 2.5 Flash API requests to this model (Gemini API pricing and deprecations pages, checked 2026-11-04).

Input tokens

  • $0.30 / 1M tokens (input) Current 2026-07-21 → present

    The same rate for text, image, video and audio input.

Cached input tokens

  • $0.03 / 1M tokens (cached input) Current 2026-07-21 → present

    Storage is billed separately at $1.00/1M tokens per hour.

Output tokens

  • $2.50 / 1M tokens (output) Current 2026-07-21 → present

    Includes thinking tokens.

Batch

Batch input tokens

  • $0.15 / 1M tokens (input) Current 2026-07-21 → present

Batch output tokens

  • $1.25 / 1M tokens (output) Current 2026-07-21 → present

Showing the active price and any recorded history. Full pricing history is available via the paid API — see API docs.

See how Gemini 3.5 Flash-Lite fits into a cost-aware routing setup

See how →

Capability profile

How Gemini 3.5 Flash-Lite rates across core capability dimensions, with the task-level evidence behind each rating.

coding Moderate

SWE-Bench Pro (Public): 54.2%; Terminal-bench 2.1 (Terminus-2 harness): 54.0%. Vendor-reported (Google DeepMind model card), in line with GPT-5.4 mini on SWE-Bench Pro (54.4%) and behind it on Terminal-bench 2.1 (59.2%).

tool use Strong

Function calling, code execution, file search and computer use all listed as supported (Gemini API model page). OSWorld-Verified (agentic computer use): 74.0%, vendor-reported (Google DeepMind model card).

context window Strong

1,048,576-token input context, 65,536-token output limit (Gemini API model page).

speed Strong

Google describes it as a "low-latency" model "optimized for high-throughput, low-cost execution" (Gemini API model page).

cost efficiency Strong

Google's models overview: "Our fastest, most cost-effective 3.5 model for high-throughput execution."

Ratings are estimated — limited independent data is available for this model.

Operator guidance

Google's low-cost Flash-Lite tier in the current Gemini 3.x line. Step up to Gemini 3.8 Flash for harder reasoning or coding; Google routes Gemini 2.5 Flash requests here from 2026-12-01.

Use cases

Limitations

Citations