// head_to_head

DeepSeek V4 Flash 0423 vs gpt-oss-20b

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

gpt-oss-20b is the cheaper of the two; neither can be ranked on quality here.

gpt-oss-20b 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.

deepseek

DeepSeek V4 Flash 0423

Blended / 1M
$0.111
Context
1.0M
Released
Apr 24, 2026
Overall score
65.5
reasoningtool callingprompt caching

openai

gpt-oss-20b

Blended / 1M
$0.036
Context
131K
Released
Aug 5, 2025
Overall score
Not evaluated
reasoningtool calling

Specs and pricing

MetricDeepSeek V4 Flash 0423gpt-oss-20b
LiveBench overall

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

65.5
Cost per point

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

$0.0083
Blended price / 1M

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

$0.111$0.036win
Input price / 1M$0.089$0.018win
Output price / 1M$0.177$0.090win
Cached input / 1M

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

$0.018
Context window1.0Mwin131K
Max output tokens384Kwin33K

Benchmarks by category

An overall score averages away the thing you probably care about. These are the seven LiveBench categories scored separately — DeepSeek V4 Flash 0423 on top, gpt-oss-20b below, both out of 100.

Agentic coding
37.6
Coding
69.2
Reasoning
70.6
Mathematics
79.6
Data analysis
68.0
Language
70.1
Instruction following
63.1

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.

WorkloadDeepSeek V4 Flash 0423gpt-oss-20b
Support chatbot

1.2K in / 400 out × 200K requests

$30.34/mo$11.52/mo
RAG assistant

8K in / 600 out × 100K requests

$53.16/mo$19.80/mo
Coding agent

40K in / 4K out × 20K requests

$45.37/mo$21.60/mo
Document extraction

20K in / 1.5K out × 50K requests

$98.35/mo$24.75/mo
Bulk classification

500 in / 20 out × 5M requests

$203.79/mo$54.00/mo
Run these two through the cost calculator

Which should you pick?

You need to fit large documents in one call

DeepSeek V4 Flash 0423

Wider context window — 1.0M against 131K.

You are cost-constrained

gpt-oss-20b

Cheaper on blended list price at $0.036 per million tokens.

DeepSeek V4 Flash 0423 vs gpt-oss-20b FAQ

Which is better, DeepSeek V4 Flash 0423 or gpt-oss-20b?

gpt-oss-20b is the cheaper of the two; neither can be ranked on quality here. gpt-oss-20b 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 DeepSeek V4 Flash 0423 cheaper than gpt-oss-20b?

gpt-oss-20b is cheaper. On a 3:1 input:output blend, DeepSeek V4 Flash 0423 lists at $0.111 per million tokens and gpt-oss-20b at $0.036 — gpt-oss-20b is 3.1× cheaper. Input and output are priced separately — DeepSeek V4 Flash 0423 charges $0.089 in and $0.177 out, gpt-oss-20b charges $0.018 and $0.090 — so the model that looks cheaper flips depending on how output-heavy your workload is.

Does DeepSeek V4 Flash 0423 or gpt-oss-20b have a bigger context window?

DeepSeek V4 Flash 0423 has the larger context window: 1.0M for DeepSeek V4 Flash 0423 against 131K for gpt-oss-20b. 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 DeepSeek V4 Flash 0423 and gpt-oss-20b support prompt caching?

DeepSeek V4 Flash 0423 publishes a cached-input rate of $0.018 per million tokens against a full input rate of $0.089. The catalogue lists no separate cached rate for gpt-oss-20b, 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.