Side-by-side comparison of pricing, 12 benchmarks, and generation speed.
| Metric | Qwen3 235B A22B 2507 (Reasoning) | GLM-4.6 (Non-reasoning) |
|---|---|---|
| Input ($/M tokens) | $0.23 | $0.57 |
| Output ($/M tokens) | $2.3 | $2.2 |
Verdict. Qwen3 235B A22B 2507 (Reasoning) and GLM-4.6 (Non-reasoning) split the benchmark comparison evenly at 1–1. The tiebreaker here is price, speed, and what you actually run them on.
Pricing. Both models sit in the budget bracket for output-token pricing. At 1.0× the per-million-token cost, GLM-4.6 (Non-reasoning) is meaningfully cheaper if your traffic is output-heavy (long completions, document generation, agent loops). GLM-4.6 (Non-reasoning) makes more sense when output volume is low and absolute reasoning quality justifies the premium.
Strengths. Qwen3 235B A22B 2507 (Reasoning) is strongest on Coding Index (22.1), Intelligence Index (19.9). GLM-4.6 (Non-reasoning) leads on Intelligence Index (23.4).
Speed. Throughput is comparable — 64 tok/s vs 62 tok/s — so generation speed shouldn't drive your choice here. Look at the per-benchmark wins instead.
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 235B A22B 2507 (Reasoning) costs $41.40 ($497/year); GLM-4.6 (Non-reasoning) costs $50.10 ($601/year). At a smaller 5M-input/2M-output scale (single-developer tool or prototype): Qwen3 235B A22B 2507 (Reasoning) ≈ $5.75/run, GLM-4.6 (Non-reasoning) ≈ $7.25/run. At agent/realtime scale (200M input / 100M output per million requests): Qwen3 235B A22B 2507 (Reasoning) ≈ $276/run, GLM-4.6 (Non-reasoning) ≈ $334/run. Qwen3 235B A22B 2507 (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
Data from Artificial Analysis API — 12 benchmarks
Qwen3 235B A22B 2507 (Reasoning) is cheaper overall. Its blended price (3:1 input/output ratio) is $0.75/M tokens vs $0.98/M for GLM-4.6 (Non-reasoning).
It's a tie — both models win 1 benchmarks each across 12 evaluated categories. See the detailed benchmark chart above for per-category results.
Qwen3 235B A22B 2507 (Reasoning) generates tokens faster at 64 tok/s vs 62 tok/s. However, GLM-4.6 (Non-reasoning) has lower time-to-first-token (2.63s vs 2.69s).
Choose based on your priorities: Qwen3 235B A22B 2507 (Reasoning) for lower cost, both perform similarly on benchmarks, and Qwen3 235B A22B 2507 (Reasoning) for faster generation. For latency-sensitive apps, check the TTFT comparison above.