// head_to_head

Gemini 3.8 Flash vs Relace Search

Published LiveBench scores across all seven categories, live list pricing, context windows, and the measured cost of a point of capability — for both models, side by side.

OpenRouter + LiveBenchAll comparisonsFull leaderboard

Gemini 3.8 Flash and Relace Search are priced within ~10% of each other.

Relace Search does not have a published LiveBench run, so this comparison covers price, context and declared capabilities only. A missing score means "not evaluated", not "worse" — the right way to separate these two is an eval on your own workload.

google

Gemini 3.8 Flash

Blended / 1M
$1.50
Context
1.0M
Released
Sep 2, 2026
Overall score
75.8
reasoningtool callingimage inputvideo inputfile inputaudio inputprompt caching

relace

Relace Search

Blended / 1M
$1.50
Context
256K
Released
Dec 8, 2025
Overall score
Not evaluated
tool calling

Specs and pricing

MetricGemini 3.8 FlashRelace Search
LiveBench overall

Mean of the seven LiveBench category scores, 0–100. Higher is better.

75.8
Cost per point

Measured benchmark spend divided by overall score — dollars per point of capability.

$0.1750
Blended price / 1M

3:1 input:output mix, the usual shape of production traffic.

$1.50$1.50
Input price / 1M$0.750win$1.00
Output price / 1M$3.75$3.00win
Cached input / 1M

Price of an input token served from the prompt cache, where the provider publishes one.

$0.075
Context window1.0Mwin256K
Max output tokens66K128Kwin

Benchmarks by category

An overall score averages away the thing you probably care about. These are the seven LiveBench categories scored separately — Gemini 3.8 Flash on top, Relace Search below, both out of 100.

Agentic coding
54.2
Coding
72.5
Reasoning
89.3
Mathematics
91.6
Data analysis
54.0
Language
87.8
Instruction following
81.4

What each one costs to run

Per-token prices are hard to feel. These are monthly list costs for both models across five workload shapes, using each provider's published cached-input rate where there is one.

WorkloadGemini 3.8 FlashRelace Search
Support chatbot

1.2K in / 400 out × 200K requests

$431.40/mo$480.00/mo
RAG assistant

8K in / 600 out × 100K requests

$555.00/mo$980.00/mo
Coding agent

40K in / 4K out × 20K requests

$522.00/mo$1,040/mo
Document extraction

20K in / 1.5K out × 50K requests

$997.50/mo$1,225/mo
Bulk classification

500 in / 20 out × 5M requests

$1,912/mo$2,800/mo
Run these two through the cost calculator

Which should you pick?

You need to fit large documents in one call

Gemini 3.8 Flash

Wider context window — 1.0M against 256K.

Gemini 3.8 Flash vs Relace Search FAQ

Which is better, Gemini 3.8 Flash or Relace Search?

Gemini 3.8 Flash and Relace Search are priced within ~10% of each other. Relace Search does not have a published LiveBench run, so this comparison covers price, context and declared capabilities only. A missing score means "not evaluated", not "worse" — the right way to separate these two is an eval on your own workload.

Is Gemini 3.8 Flash cheaper than Relace Search?

They cost about the same. Both land near $1.50 per million tokens on a 3:1 input:output blend, so price is unlikely to be the deciding factor between them.

Does Gemini 3.8 Flash or Relace Search have a bigger context window?

Gemini 3.8 Flash has the larger context window: 1.0M for Gemini 3.8 Flash against 256K for Relace Search. Note that a window you can fill is not a window you should fill — retrieval quality usually degrades well before the limit, and you pay for every token you put in it.

Do Gemini 3.8 Flash and Relace Search support prompt caching?

Gemini 3.8 Flash publishes a cached-input rate of $0.075 per million tokens against a full input rate of $0.750. The catalogue lists no separate cached rate for Relace Search, which means the provider does not price it separately here — not that caching is unavailable.

Related comparisons

How these numbers are produced

  • Price — provider list price from OpenRouter, refreshed every 15 minutes. “Blended” is a 3:1 input:output mix.
  • ScoresLiveBench release 2026-06-25, using their own category map. Each model shows its strongest published run. A blank means “not evaluated”, never “bad”.
  • Cost per point — the measured dollars LiveBench spent on the run, divided by the score it earned.
  • “Win” — awarded only past a threshold: one full point on a benchmark score, 10% on a price, 25% on a context window. Anything tighter reports as a tie, because effort settings alone move a LiveBench score by more than that.

Published benchmarks rank models on someone else's tasks. Before committing, see LLM & agent evaluation for building an eval on your own.