Compare/Qwen3.6 27B (Reasoning) vs GLM-5 (Non-reasoning)

Qwen3.6 27B (Reasoning)vsGLM-5 (Non-reasoning)

Side-by-side comparison of pricing, 12 benchmarks, and generation speed.

Alibaba

Qwen3.6 27B (Reasoning)

Input
$0.6/M
Output
$3.6/M
Speed
58 tok/s
TTFT
1.43s
Z AI

GLM-5 (Non-reasoning)

Input
$1/M
Output
$3.2/M
Speed
TTFT

Winner by Category

Cheaper
Qwen3.6 27B (Reasoning)
Faster (tok/s)
Qwen3.6 27B (Reasoning)
Lower Latency
GLM-5 (Non-reasoning)
Benchmarks (6-1)
Qwen3.6 27B (Reasoning)

Pricing Comparison

MetricQwen3.6 27B (Reasoning)GLM-5 (Non-reasoning)
Input ($/M tokens)$0.6$1
Output ($/M tokens)$3.6$3.2
Cost for 1M input + 100K output tokens:
Qwen3.6 27B (Reasoning)$0.96
GLM-5 (Non-reasoning)$1.32

Speed Comparison

Output Speed (tokens/s) — higher is better
Qwen3.6 27B (Reasoning)
58 tok/s
GLM-5 (Non-reasoning)
Time to First Token (seconds) — lower is better
Qwen3.6 27B (Reasoning)
1.43s
GLM-5 (Non-reasoning)

Editorial Analysis

Verdict. Qwen3.6 27B (Reasoning) wins the overall benchmark matchup 6–1 across 7 overlapping categories, but raw benchmark score is only one input to the decision.

Pricing. Both models sit in the budget bracket for output-token pricing. At 1.1× the per-million-token cost, GLM-5 (Non-reasoning) is meaningfully cheaper if your traffic is output-heavy (long completions, document generation, agent loops). GLM-5 (Non-reasoning) makes more sense when output volume is low and absolute reasoning quality justifies the premium.

Strengths. Qwen3.6 27B (Reasoning) is strongest on GPQA Diamond (84%), IFBench (68%), Coding Index (53.7). GLM-5 (Non-reasoning) leads on GPQA Diamond (67%), IFBench (55%), TerminalBench (39%).

Speed. On throughput, Qwen3.6 27B (Reasoning) generates tokens at 58 tok/s versus 0 tok/s — about 100% faster. On time-to-first-token, GLM-5 (Non-reasoning) responds in 0ms vs 1429ms, which matters most for chat-style UIs.

Provider. Alibaba and Z AI sell to overlapping but distinct developer audiences: Alibaba tends to ship frontier reasoning models with premium positioning, while Z AI often prices more aggressively. Your existing vendor relationships, billing, and SLA preferences may matter as much as the raw numbers above.

Workload cost. Workload scenarios (per million requests at 30M input + 15M output tokens): Qwen3.6 27B (Reasoning) costs $72.00 ($864/year); GLM-5 (Non-reasoning) costs $78.00 ($936/year). At a smaller 5M-input/2M-output scale (single-developer tool or prototype): Qwen3.6 27B (Reasoning) ≈ $10.20/run, GLM-5 (Non-reasoning) ≈ $11.40/run. At agent/realtime scale (200M input / 100M output per million requests): Qwen3.6 27B (Reasoning) ≈ $480/run, GLM-5 (Non-reasoning) ≈ $520/run. Qwen3.6 27B (Reasoning) becomes more attractive at higher volume — the absolute per-token pricing difference compounds when you ship at scale.

Recommendation. Both models have legitimate use cases — the right answer depends on whether you are optimizing for benchmark ceiling, latency, or unit cost. Start with the cheaper / faster model, evaluate against your specific task, and only switch if the upgrade shows a meaningful lift.

Head-to-head deltas

  • Qwen3.6 27B (Reasoning) wins 5 more benchmarks than its opponent — a margin wide enough to call the comparison settled on benchmark terms alone.
  • On throughput, Qwen3.6 27B (Reasoning) is 5842.70× faster (58 tok/s vs 0 tok/s). For streaming chat or real-time agents this alone often flips the recommendation.
  • Time-to-first-token differs by 14290.0× — GLM-5 (Non-reasoning) responds in 0ms vs 1429ms. For interactive chat UIs this can matter more than raw benchmark wins.

Benchmark Comparison

Data from Artificial Analysis API — 12 benchmarks

Intelligence Index
37.132.4
Coding Index
53.7
Math Index
GPQA Diamond
84.2%66.6%
MMLU-Pro
LiveCodeBench
AIME 2025
MATH-500
Humanity's Last Exam
21.6%7.2%
SciCode
39.8%38.3%
IFBench
67.6%55.2%
TerminalBench
34.8%39.4%
Qwen3.6 27B (Reasoning)6 wins
1 winsGLM-5 (Non-reasoning)

Frequently Asked Questions

Which is cheaper, Qwen3.6 27B (Reasoning) or GLM-5 (Non-reasoning)?

Qwen3.6 27B (Reasoning) is cheaper overall. Its blended price (3:1 input/output ratio) is $1.35/M tokens vs $1.55/M for GLM-5 (Non-reasoning).

Which model performs better on benchmarks?

Qwen3.6 27B (Reasoning) wins 6 out of 12 benchmarks compared to 1 for GLM-5 (Non-reasoning). See the detailed benchmark chart above for per-category results.

Which is faster for real-time applications?

Qwen3.6 27B (Reasoning) generates tokens faster at 58 tok/s vs 0 tok/s. However, GLM-5 (Non-reasoning) has lower time-to-first-token (0.00s vs 1.43s).

When should I use Qwen3.6 27B (Reasoning) vs GLM-5 (Non-reasoning)?

Choose based on your priorities: Qwen3.6 27B (Reasoning) for lower cost, Qwen3.6 27B (Reasoning) for stronger benchmark performance, and Qwen3.6 27B (Reasoning) for faster generation. For latency-sensitive apps, check the TTFT comparison above.