// head_to_head

Qwen3.6 Plus vs Qwen3.7 Plus

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

Qwen3.7 Plus is the cheaper of the two; neither can be ranked on quality here.

Qwen3.7 Plus 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.

qwen

Qwen3.6 Plus

Blended / 1M
$0.731
Context
1M
Released
Apr 2, 2026
Overall score
68.9
reasoningtool callingimage inputvideo input

qwen

Qwen3.7 Plus

Blended / 1M
$0.560
Context
1M
Released
Jun 3, 2026
Overall score
Not evaluated
reasoningtool callingimage inputprompt caching

Specs and pricing

MetricQwen3.6 PlusQwen3.7 Plus
LiveBench overall

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

68.9
Cost per point

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

$0.1262
Blended price / 1M

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

$0.731$0.560win
Input price / 1M$0.325$0.320
Output price / 1M$1.95$1.28win
Cached input / 1M

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

$0.064
Context window1M1M
Max output tokens66K131Kwin

Benchmarks by category

An overall score averages away the thing you probably care about. These are the seven LiveBench categories scored separately — Qwen3.6 Plus on top, Qwen3.7 Plus below, both out of 100.

Agentic coding
41.4
Coding
78.2
Reasoning
75.8
Mathematics
83.7
Data analysis
69.9
Language
75.0
Instruction following
58.3

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.

WorkloadQwen3.6 PlusQwen3.7 Plus
Support chatbot

1.2K in / 400 out × 200K requests

$234.00/mo$160.77/mo
RAG assistant

8K in / 600 out × 100K requests

$377.00/mo$230.40/mo
Coding agent

40K in / 4K out × 20K requests

$416.00/mo$215.04/mo
Document extraction

20K in / 1.5K out × 50K requests

$471.25/mo$403.20/mo
Bulk classification

500 in / 20 out × 5M requests

$1,008/mo$800.00/mo
Run these two through the cost calculator

Which should you pick?

You are cost-constrained

Qwen3.7 Plus

Cheaper on blended list price at $0.560 per million tokens.

Qwen3.6 Plus vs Qwen3.7 Plus FAQ

Which is better, Qwen3.6 Plus or Qwen3.7 Plus?

Qwen3.7 Plus is the cheaper of the two; neither can be ranked on quality here. Qwen3.7 Plus 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 Qwen3.6 Plus cheaper than Qwen3.7 Plus?

Qwen3.7 Plus is cheaper. On a 3:1 input:output blend, Qwen3.6 Plus lists at $0.731 per million tokens and Qwen3.7 Plus at $0.560 — Qwen3.7 Plus is 31% cheaper. Input and output are priced separately — Qwen3.6 Plus charges $0.325 in and $1.95 out, Qwen3.7 Plus charges $0.320 and $1.28 — so the model that looks cheaper flips depending on how output-heavy your workload is.

Does Qwen3.6 Plus or Qwen3.7 Plus have a bigger context window?

They are effectively the same — 1M for Qwen3.6 Plus and 1M for Qwen3.7 Plus.

Do Qwen3.6 Plus and Qwen3.7 Plus support prompt caching?

Qwen3.7 Plus publishes a cached-input rate of $0.064 per million tokens against a full input rate of $0.320. The catalogue lists no separate cached rate for Qwen3.6 Plus, 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.