Strengths
- One of the fastest models: 460.6 tokens per second, third of 149 in Artificial Analysis measurements.
- Google says it beats the larger Gemini 3 Flash on the SWE-Bench Pro coding test (54.2% vs 49.6%) and on OSWorld-Verified computer use (74.0% vs 65.1%).
- A low price of $0.30 input and $2.50 output per million tokens, plus a one-million-token context and input from images, video, audio and PDFs. You can also try it in the API for free.
- It keeps answers short. It produced 44 million output tokens on the Artificial Analysis test, the seventh most economical of 22 in its group.
Weaknesses
- Not suited to hard reasoning: it scores 12 on the Artificial Analysis index, 19th of 22.
- Output at $2.50 per million tokens is noticeably pricier than the older 3.1 Flash-Lite ($1.50) and 2.5 Flash-Lite ($0.40).
- For the free Gemini app, Google only says “Flash-Lite” without a version number, so it is not certain this exact model runs there.
Best for
- processing large numbers of documents and extracting data
- sorting requests and helper tasks for larger agents
- fast search tasks where response time matters
Not ideal for
- complex reasoning and difficult coding, where Gemini 3.8 Flash or Gemini 4 Argon fit better
- long expert writing
How to use it in Europe
Since 9 October 2026, the Gemini app without a subscription sends most questions to a Flash-Lite model with a 32K-token context. Google does not say which version. The app works in Slovakia and Czechia, in Slovak and Czech too. Developers use the model through the Gemini API and Google AI Studio, on both the free and paid tiers. Under its terms, Google does not train on API data from users in the EEA, even on the free tier.
Price compared with similar models
Benchmark results
| Benchmark | Score | Source |
|---|---|---|
| SWE-Bench Pro (Google) | 54.2% | blog.google ↗, July 2026 |
| OSWorld-Verified (Google) | 74.0% | blog.google ↗, July 2026 |
| Terminal-Bench 2.1 (Google) | 54% | blog.google ↗, July 2026 |
| Artificial Analysis output speed | 460.6 tokens/s (#3 of 149) | artificialanalysis.ai ↗, October 2026 |
| Artificial Analysis Intelligence Index | 12 (#19 of 22) | artificialanalysis.ai ↗, October 2026 |
Benchmarks cover only a narrow slice of ability and vendors often report them at their best settings. Treat them as a rough guide.
More models from this vendor
- Gemini 4 Carbon: Google's most capable model, announced on 28 October 2026. So far only selected cybersecurity experts have access.
- Gemini 4 Argon: Google's most capable publicly available model, out in October 2026. Built for demanding analytical work where accuracy matters and made-up answers are costly.
- Gemini 3.8 Flash: Google's fast, lower-cost model for everyday work, coding and document analysis. You can try it in the API for free.
- Gemma 4 31B: Google's largest open model, which you can run on your own computer or server when data must not leave the company.
- Gemma 4 26B A4B: Google's faster open model for your own server. Close to the larger Gemma 4 31B in quality, but quicker to respond.
Similar models from other vendors
- Claude Haiku 5.5 (Anthropic): Claude's cheapest and fastest model, built for classification, data extraction and large volumes of simple tasks.
- DeepSeek V4.1 Flash (DeepSeek): DeepSeek's cheap, fast open-weight model with a 1 million token window. In the app and API, though, your data goes to China.
- Mistral Small 4 (Mistral AI): Mistral's cheap, fast open model under Apache 2.0 for everyday tasks through the API or on your own server.
- GPT-6.1 Terra (OpenAI): OpenAI's fast, cheaper model for working with images, operating a computer and everyday office work.
- Qwen3.8-Flash (Alibaba (Qwen)): A very cheap, fast Qwen model for high volumes of tasks. Unlike Max and Plus, it runs in Frankfurt with processing kept in the EU.
- Grok 4.3 (xAI): An older, cheaper Grok from April 2026 with a 1-million-token window. Good for lots of low-cost API calls.