// head_to_head

Qwen3.6 27B vs Llama 3.1 Euryale 70B v2.2

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

Qwen3.6 27B and Llama 3.1 Euryale 70B v2.2 are priced within ~10% of each other.

Llama 3.1 Euryale 70B v2.2 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 27B

Blended / 1M
$0.915
Context
262K
Released
Apr 27, 2026
Overall score
64.0
reasoningtool callingimage inputvideo inputprompt caching

sao10k

Llama 3.1 Euryale 70B v2.2

Blended / 1M
$0.850
Context
131K
Released
Aug 28, 2024
Overall score
Not evaluated
tool calling

Specs and pricing

MetricQwen3.6 27BLlama 3.1 Euryale 70B v2.2
LiveBench overall

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

64.0
Cost per point

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

$0.1074
Blended price / 1M

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

$0.915$0.850
Input price / 1M$0.320win$0.850
Output price / 1M$2.70$0.850win
Cached input / 1M

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

$0.150
Context window262Kwin131K
Max output tokens262Kwin16K

Benchmarks by category

An overall score averages away the thing you probably care about. These are the seven LiveBench categories scored separately — Qwen3.6 27B on top, Llama 3.1 Euryale 70B v2.2 below, both out of 100.

Agentic coding
39.3
Coding
71.8
Reasoning
70.3
Mathematics
79.9
Data analysis
70.4
Language
63.3
Instruction following
53.2

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 27BLlama 3.1 Euryale 70B v2.2
Support chatbot

1.2K in / 400 out × 200K requests

$280.56/mo$272.00/mo
RAG assistant

8K in / 600 out × 100K requests

$350.00/mo$731.00/mo
Coding agent

40K in / 4K out × 20K requests

$376.80/mo$748.00/mo
Document extraction

20K in / 1.5K out × 50K requests

$514.00/mo$913.75/mo
Bulk classification

500 in / 20 out × 5M requests

$985.00/mo$2,210/mo
Run these two through the cost calculator

Which should you pick?

You need to fit large documents in one call

Qwen3.6 27B

Wider context window — 262K against 131K.

Qwen3.6 27B vs Llama 3.1 Euryale 70B v2.2 FAQ

Which is better, Qwen3.6 27B or Llama 3.1 Euryale 70B v2.2?

Qwen3.6 27B and Llama 3.1 Euryale 70B v2.2 are priced within ~10% of each other. Llama 3.1 Euryale 70B v2.2 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 27B cheaper than Llama 3.1 Euryale 70B v2.2?

They cost about the same. Both land near $0.915 per million tokens on a 3:1 input:output blend, so price is unlikely to be the deciding factor between them.

Does Qwen3.6 27B or Llama 3.1 Euryale 70B v2.2 have a bigger context window?

Qwen3.6 27B has the larger context window: 262K for Qwen3.6 27B against 131K for Llama 3.1 Euryale 70B v2.2. 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 Qwen3.6 27B and Llama 3.1 Euryale 70B v2.2 support prompt caching?

Qwen3.6 27B publishes a cached-input rate of $0.150 per million tokens against a full input rate of $0.320. The catalogue lists no separate cached rate for Llama 3.1 Euryale 70B v2.2, 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.