DeepSeekDSH
Independent community guideNot affiliated with DeepSeek.Download versions and guide references

Choose a Gemini Model by Task and Cost

Beginner guide to Google Gemini model selection: compare 3.8 Flash, 4 Argon, and DeepSeek prices, watch 2027 rate changes, and set up DSH in 4 simple steps.

Maintained by DeepSeekDSH (independent site)Documentation review
On this page
Gemini model selection illustration routing document and coding tasks

Opening Google's Gemini model list for the first time can feel overwhelming when you see Flash, Pro, Argon, Live, and TTS stacked together.

You do not need to memorize version history. Match the job on your desk to one plain rule:

  • For 90% of everyday work (reading PDFs, images, audio, video, and normal coding): start with fast, affordable Gemini 3.8 Flash.
  • For tough multi-file code refactors and deep debugging: step up to Google's heavy engineering model Gemini 4 Argon.
  • For live two-way voice chat or text-to-speech audio files: use the dedicated Live or TTS audio models.

Side-by-side comparison: Google Gemini models vs DeepSeek

Combining Google's official model directory and official Gemini API pricing, here is how the main options line up (prices in US dollars per 1 million tokens, roughly 750,000 English words):

Model NameModel IDAccepted File InputsInput / Output Price (per 1M tokens)Plain-English Fit
Gemini 3.8 Flashgemini-3.8-flashText, images, PDFs, audio, and video$0.75 / $3.75 ($1.50 / $7.50 starting Jan 1, 2027)Best everyday all-rounder: fast and cheap for PDFs, screenshots, and routine coding
Gemini 4 Argongemini-4-argonLarge codebases and multimodal context$2.00 / $10.00 intro ($4 / $20 from Dec 2026)Heavy coding specialist: built for multi-file refactors and long patches
Gemini Live / TTSDedicated audio IDsReal-time voice streams / speech outputMetered on audio/speech tiersVoice bots and spoken narration only—not for normal code editing
DeepSeek V4.1 Flash (comparison)deepseek-flashText and code focused$0.30 / $1.20 (50% off-peak)Cheaper alternative when your task is pure text or code without media files

Still unsure? Filter in three quick steps:

  1. Check your input files: If you need to feed in a PDF invoice, audio recording, or screenshot, pick Gemini 3.8 Flash (full breakdown in our 3.8 Flash analysis).
  2. Pure text or code on a tight budget? Compare Gemini 3.8 Flash against DeepSeek V4.1 Flash (side-by-side math in DeepSeek vs Gemini 3.8 Flash).
  3. Stuck on a complex multi-file bug? Escalate just that tricky session to Gemini 4 Argon instead of paying top-tier prices all day.

Real dollar example: what does one Gemini 3.8 Flash prompt cost?

At promotional rates (valid through December 31, 2026), suppose you send 10,000 input tokens (~7,500 words of docs and code) and Gemini writes back 2,000 output tokens (~1,500 words):

  • Input cost: 0.01 × $0.75 = $0.0075
  • Output cost: 0.002 × $3.75 = $0.0075
  • Total for that turn: $0.015 (one and a half cents).

Keep two billing details in mind when planning ahead:

  1. Watch the promo end date: On January 1, 2027, Gemini 3.8 Flash moves from $0.75 / $3.75 to its regular $1.50 / $7.50 list rate.
  2. Thinking tokens count toward output: Google bills internal reasoning (thinking tokens) at the output rate, so higher reasoning settings use more billed output tokens than the visible reply alone.

4-step beginner setup: connect Google Gemini in DSH

Follow these four steps to connect Gemini inside DeepSeek Harness (DSH) without hitting the common custom-provider trap:

Step 1: Grab a Gemini API key from Google AI Studio

Open aistudio.google.com/apikey in your browser (details in Google's official API key guide), sign in with your Google account, click Create API key, and copy the AIza... key into a private note.

Step 2: Pick the built-in "Google" card in DSH (do NOT use Custom Provider!)

Launch DSH on your computer and click Settings -> Models -> Add provider:

  • Right way: Click the built-in Google card, paste your AIza... key, and click Save (full walkthrough in our DSH Google provider guide).
  • Beginner trap to avoid: Do not click Add a custom provider for native Gemini keys! DSH's built-in Google card speaks Google's native google-generative-ai protocol, whereas the custom provider form only speaks OpenAI and Anthropic protocols.

Step 3: Verify in an empty test folder

Create an empty dsh-test folder on your Desktop, click Choose workspace in DSH to select it, pick gemini-3.8-flash from the model dropdown, and send "Reply only with connected; do not edit files".

Step 4: One parameter tip if you also write Python scripts

If you call gemini-3.8-flash from Python (for example, following DataCamp's Gemini 3.8 Flash tutorial for PDF extraction), remember that 3.8 Flash accepts low, medium, and high reasoning levels—minimal is unsupported, so avoid copying legacy Flash settings verbatim.

Common questions

Is it safe to paste private company code on Google AI Studio's Free tier?

No—check your billing tier first! Google's Free tier terms allow prompts to be used for product improvement, whereas paid Pay-as-you-go projects have separate commercial data-privacy protections. Always verify your AI Studio project's billing status before sending confidential files.

Why does my key work for 3.8 Flash yet fail on Gemini 4 Argon?

Gemini 4 Argon is not included on the Gemini API Free tier. Enable Pay-as-you-go billing on your Google AI Studio / Cloud project first, and the same project can call Argon.

Is the most expensive Pro or Argon model always better for everyday tasks?

Rarely. Simple scripts, formatting fixes, and document extraction run faster and several times cheaper on Gemini 3.8 Flash. Save heavier tiers for tasks where Flash actually misses a requirement.

Model and workflow guides

Models and costs

Gemini 4 Neon is here. When is it worth switching models?

Compare Gemini 4 Neon in two coding setups, work through a $0.45 API example, and find Neon in Gemini, Google AI Studio, and Antigravity.

Read →
Models and costs

GPT-6.2 Astra & Sol Guide: Specs, Pricing & Choice

Beginner guide to OpenAI's GPT-6.2 Astra and Sol: Claude vs DeepSeek price table, plain-English benchmarks, 272K cliff, and step-by-step DSH & Python setup.

Read →
Models and costs

GPT-6.1 Terra Guide: Uses, Pricing and Sol Comparison

Beginner guide to OpenAI's GPT-6.1 Terra: half-price visual, computer-use, and writing model compared with Sol and DeepSeek, plus DSH and Python setup steps.

Read →
Model setup and costs

Connect Local Qwen 3.8 27B to DSH

Beginner guide to connecting local Qwen 3.8 27B to DSH: GPU memory table, LM Studio / Ollama / llama.cpp port setup, exact 5-field config, and error fixes.

Read →
Model setup and costs

Codex Costs: API Billing vs Subscription Usage

Beginner guide to OpenAI Codex costs: ChatGPT Plus login vs API-key billing table, real per-prompt dollar math, and 3 steps to lock a monthly budget cap.

Read →
Model setup and costs

Connect Grok API to DSH: Setup & Tools

Step-by-step beginner guide to connecting xAI's Grok API (grok-4.7) to DSH: Grok vs Groq comparison, exact 5-field custom provider setup, and 401/404 fixes.

Read →
Model benchmarks and engineering

Gemini 4 Argon Coding Benchmarks & 1M Window

Beginner guide to Google's Gemini 4 Argon coding scores vs Claude and GPT, how its shared 1M window and 65K output cap work, plus 4-step DSH setup.

Read →
API costs and enterprise ROI

Gemini 4 Argon API Pricing, Cache & ROI Guide

Beginner guide to Gemini 4 Argon API pricing: Dec 2026 2x rate jump, how 95% auto-caching and thinking tokens bill, DeepSeek cost comparison, and calculator.

Read →
Model setup and costs

Claude Sonnet vs Opus: Which Should You Use?

Beginner guide to Claude Sonnet 5.5 vs Opus 5.5: side-by-side API prices, per-prompt cost math, clear rules on when to upgrade, and 4-step DSH setup.

Read →
Models and costs

Opus 5.5 vs Fable 5.2: Which Model Is Right for You?

Beginner guide to Claude Opus 5.5 vs Fable 5.2: side-by-side API prices, coding benchmarks, per-task cost math, 30-day retention rules, and 4-step DSH setup.

Read →
Content creation

How to draft a video script with Media Desk for DeepSeek Harness

Turn a phone note into a video script, cover, and storyboard with Media Desk for DeepSeek Harness. See how to install it, set up images, and export a draft.

Read →
Mobile and remote access

Use DeepSeek Harness on Your Phone: Beginner Guide

Zero-jargon beginner guide to using DeepSeek Harness on your phone: find your computer IP, connect over home Wi-Fi in 2 minutes, or use Tailscale on 4G/5G.

Read →
Model interfaces and workflows

MiniMax H3 vs M-Series in DSH: Which Works?

Beginner guide to MiniMax H3 vs M3 in DSH: why H3 is for video while M3 handles coding, how to avoid region 401 errors, and 4-step DSH setup instructions.

Read →
API credentials and troubleshooting

Kimi API Key Errors: Region, Model and Permissions

Beginner fix guide for Kimi / Moonshot API key errors: match .ai vs .cn URLs, avoid retired model names and K3 Max typos, decode 401/404/429, and set up DSH.

Read →
Image editing and tools

AI Image Editing: Prompts, Masks and Review

Beginner guide to AI image editing without ruining your subject: copy our 4-line prompt template, avoid PNG mask traps, inspect edges, and use DSH safely.

Read →