Doubao 1.5 Pro Pro vs GLM-4.6: which model should you choose?
We compared Doubao 1.5 Pro Pro and GLM-4.6 on price, context window and 12 benchmarks. Doubao 1.5 Pro Pro and GLM-4.6 are effectively tied on the index average; GLM-4.6 is cheaper on a blended 3:1 basis.
Doubao 1.5 Pro Pro
Doubao 1.5 Pro Pro is a multimodal model from ByteDance. It offers a 128,000-token context window, costs $14.10 per million input tokens and $54.74 per million output tokens, and averages 80.2 across the index's benchmark suite.
- Input
- $14.10 / 1M
- Output
- $54.74 / 1M
- Context
- 128,000
- Avg score
- 80.2
- Open weights
- No
GLM-4.6
GLM-4.6 is a multimodal model from Zhipu AI. It offers a 128,000-token context window, costs $5.10 per million input tokens and $18.45 per million output tokens, and averages 80.2 across the index's benchmark suite.
- Input
- $5.10 / 1M
- Output
- $18.45 / 1M
- Context
- 128,000
- Avg score
- 80.2
- Open weights
- Yes
Specification comparison
| Spec | Doubao 1.5 Pro Pro | GLM-4.6 | Edge |
|---|---|---|---|
| Input price / 1M | $14.10 | $5.10 | B |
| Output price / 1M | $54.74 | $18.45 | B |
| Context window | 128,000 | 128,000 | Tie |
| Max output | 8,192 | 8,192 | Tie |
| Average score | 80.2 | 80.2 | Tie |
| Arena Elo | 1443.4 | 1440.7 | A |
| Open weights | No | Yes | B |
Benchmark comparison
Doubao 1.5 Pro Pro wins 7, GLM-4.6 wins 4, and 1 benchmarks are tied. The table shows the modelled score for each model and the gap.
| Benchmark | Doubao 1.5 Pro Pro | GLM-4.6 | Gap | Winner |
|---|---|---|---|---|
| MMLU-Pro | 80.2 | 84.0 | 3.8 | GLM-4.6 |
| GPQA Diamond | 74.5 | 78.3 | 3.8 | GLM-4.6 |
| SWE-bench Verified | 71.2 | 67.4 | 3.8 | Doubao 1.5 Pro Pro |
| HumanEval | 84.4 | 83.8 | 0.6 | Doubao 1.5 Pro Pro |
| MATH-500 | 85.4 | 83.4 | 2.0 | Doubao 1.5 Pro Pro |
| AIME 2025 | 77.7 | 80.2 | 2.5 | GLM-4.6 |
| LiveCodeBench | 71.6 | 71.6 | 0.0 | Tie |
| IFEval | 87.8 | 86.4 | 1.4 | Doubao 1.5 Pro Pro |
| TruthfulQA | 84.6 | 82.5 | 2.1 | Doubao 1.5 Pro Pro |
| HellaSwag | 87.2 | 89.4 | 2.2 | GLM-4.6 |
| MMMU | 78.1 | 75.4 | 2.7 | Doubao 1.5 Pro Pro |
| Arena Elo | 1443.4 | 1440.7 | 2.7 | Doubao 1.5 Pro Pro |
Price analysis
On a blended 3:1 input-to-output basis, Doubao 1.5 Pro Pro costs about $24.26 per million tokens and GLM-4.6 costs about $8.44. GLM-4.6 is roughly 65% cheaper than Doubao 1.5 Pro Pro. At high volume, that gap compounds quickly, so price should carry real weight unless the quality difference is decisive.
Both models list a 128,000-token context window. That is enough for long documents, large codebases and multi-step agent traces.
Verdict
Across 12 shared benchmarks, Doubao 1.5 Pro Pro wins 7, GLM-4.6 wins 4, and 1 are tied. The two models are effectively tied on the index average at 80.2.
The price picture favours GLM-4.6. The context picture is effectively a tie. Those two dimensions often matter more than a few benchmark points once a model is good enough for the task.
If you need a single default, start with GLM-4.6 if cost is the constraint and Doubao 1.5 Pro Pro if context is the constraint. Test both on a sample of your own traffic before committing.
Choose Doubao 1.5 Pro Pro if
- You prefer ByteDance's tooling or already use it.
Choose GLM-4.6 if
- Unit cost matters and you want the cheaper blended price.
- You need open weights for self-hosting or fine-tuning.
FAQ
Is Doubao 1.5 Pro Pro better than GLM-4.6?
The two models are effectively tied on the index average at 80.2. The better choice depends on price, context and your task. Doubao 1.5 Pro Pro wins 7 benchmarks, GLM-4.6 wins 4.
Which is cheaper, Doubao 1.5 Pro Pro or GLM-4.6?
GLM-4.6 is cheaper on a blended 3:1 basis at about $8.44 per million tokens, versus $24.26 for Doubao 1.5 Pro Pro.
Which has the larger context window?
Both models list a 128,000-token context window.
Are these models open weights?
Doubao 1.5 Pro Pro is proprietary; GLM-4.6 is open weights.
Which should I choose for production?
If you need a single default, start with GLM-4.6 if cost is the constraint and Doubao 1.5 Pro Pro if context is the constraint. Test both on a sample of your own traffic before committing.