Model Comparison

DeepSeek-V3.2 (Thinking) vs GLM-4.7

GLM-4.7 shows notably better performance in the majority of benchmarks. DeepSeek-V3.2 (Thinking) is 3.2x cheaper per token.

Want to compare interactively?Try the playground

Performance Benchmarks

Comparative analysis across standard metrics

9 benchmarks

DeepSeek-V3.2 (Thinking) outperforms in 3 benchmarks (MMLU-Pro, SWE-bench Multilingual, Terminal-Bench 2.0), while GLM-4.7 is better at 6 benchmarks (AIME 2025, BrowseComp, BrowseComp-zh, GPQA, Humanity's Last Exam, SWE-Bench Verified).

GLM-4.7 shows notably better performance in the majority of benchmarks.

Mon May 18 2026 • llm-stats.com

Arena Performance

Human preference votes

Pricing Analysis

Price comparison per million tokens

DeepSeek-V3.2 (Thinking) costs less

For input processing, DeepSeek-V3.2 (Thinking) ($0.28/1M tokens) is 2.1x cheaper than GLM-4.7 ($0.60/1M tokens).

For output processing, DeepSeek-V3.2 (Thinking) ($0.42/1M tokens) is 5.2x cheaper than GLM-4.7 ($2.20/1M tokens).

In conclusion, GLM-4.7 is more expensive than DeepSeek-V3.2 (Thinking).*

* Using a 3:1 ratio of input to output tokens

Lowest available price from all providers
Mon May 18 2026 • llm-stats.com
DeepSeek
DeepSeek-V3.2 (Thinking)
Input tokens$0.28
Output tokens$0.42
Best providerDeepSeek
Zhipu AI
GLM-4.7
Input tokens$0.60
Output tokens$2.20
Best providerFireworks
Notice missing or incorrect data?Start an Issue→

Model Size

Parameter count comparison

327.0B diff

DeepSeek-V3.2 (Thinking) has 327.0B more parameters than GLM-4.7, making it 91.3% larger.

DeepSeek
DeepSeek-V3.2 (Thinking)
685.0Bparameters
Zhipu AI
GLM-4.7
358.0Bparameters
685.0B
DeepSeek-V3.2 (Thinking)
358.0B
GLM-4.7

Context Window

Maximum input and output token capacity

GLM-4.7 accepts 202,800 input tokens compared to DeepSeek-V3.2 (Thinking)'s 131,072 tokens. GLM-4.7 can generate longer responses up to 131,072 tokens, while DeepSeek-V3.2 (Thinking) is limited to 65,536 tokens.

DeepSeek
DeepSeek-V3.2 (Thinking)
Input131,072 tokens
Output65,536 tokens
Zhipu AI
GLM-4.7
Input202,800 tokens
Output131,072 tokens
Mon May 18 2026 • llm-stats.com

Input Capabilities

Supported data types and modalities

GLM-4.7 supports multimodal inputs, whereas DeepSeek-V3.2 (Thinking) does not.

GLM-4.7 can handle both text and other forms of data like images, making it suitable for multimodal applications.

DeepSeek-V3.2 (Thinking)

Text
Images
Audio
Video

GLM-4.7

Text
Images
Audio
Video

License

Usage and distribution terms

Both models are licensed under MIT.

Both models share the same licensing terms, providing consistent usage rights.

DeepSeek-V3.2 (Thinking)

MIT

Open weights

GLM-4.7

MIT

Open weights

Release Timeline

When each model was launched

DeepSeek-V3.2 (Thinking) was released on 2025-12-01, while GLM-4.7 was released on 2025-12-22.

GLM-4.7 is 1 month newer than DeepSeek-V3.2 (Thinking).

DeepSeek-V3.2 (Thinking)

Dec 1, 2025

5 months ago

GLM-4.7

Dec 22, 2025

4 months ago

3w newer

Knowledge Cutoff

When training data ends

Neither model specifies a knowledge cutoff date.

Unable to compare the recency of their training data.

No cutoff dates available

Provider Availability

DeepSeek-V3.2 (Thinking) is available from DeepSeek. GLM-4.7 is available from Fireworks, Novita.

DeepSeek-V3.2 (Thinking)

deepseek logo
DeepSeek
Input Price:Input: $0.28/1MOutput Price:Output: $0.42/1M

GLM-4.7

fireworks logo
Fireworks
Input Price:Input: $0.60/1MOutput Price:Output: $2.20/1M
novita logo
Novita
Input Price:Input: $0.60/1MOutput Price:Output: $2.20/1M
* Prices shown are per million tokens

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion→

Key Takeaways

DeepSeek
Less expensive input tokens
Less expensive output tokens
Higher MMLU-Pro score (85.0% vs 84.3%)
Higher SWE-bench Multilingual score (70.2% vs 66.7%)
Higher Terminal-Bench 2.0 score (46.4% vs 41.0%)
Zhipu AI
Larger context window (202,800 tokens)
Supports multimodal inputs
Higher AIME 2025 score (95.7% vs 93.1%)
Higher BrowseComp score (52.0% vs 51.4%)
Higher BrowseComp-zh score (66.6% vs 65.0%)
Higher GPQA score (85.7% vs 82.4%)
Higher Humanity's Last Exam score (42.8% vs 25.1%)
Higher SWE-Bench Verified score (73.8% vs 73.1%)

Detailed Comparison

AI Model Comparison Table
Feature
DeepSeek
DeepSeek-V3.2 (Thinking)
Zhipu AI
GLM-4.7

FAQ

Common questions about DeepSeek-V3.2 (Thinking) vs GLM-4.7.

Which is better, DeepSeek-V3.2 (Thinking) or GLM-4.7?

GLM-4.7 shows notably better performance in the majority of benchmarks. DeepSeek-V3.2 (Thinking) is made by DeepSeek and GLM-4.7 is made by Zhipu AI. The best choice depends on your use case — compare their benchmark scores, pricing, and capabilities above.

How does DeepSeek-V3.2 (Thinking) compare to GLM-4.7 in benchmarks?

DeepSeek-V3.2 (Thinking) scores AIME 2025: 93.1%, HMMT 2025: 90.2%, MMLU-Pro: 85.0%, LiveCodeBench: 83.3%, GPQA: 82.4%. GLM-4.7 scores AIME 2025: 95.7%, Tau-bench: 87.4%, GPQA: 85.7%, LiveCodeBench v6: 84.9%, MMLU-Pro: 84.3%.

Is DeepSeek-V3.2 (Thinking) cheaper than GLM-4.7?

DeepSeek-V3.2 (Thinking) is 2.1x cheaper for input tokens. DeepSeek-V3.2 (Thinking) costs $0.28/M input and $0.42/M output via deepseek. GLM-4.7 costs $0.60/M input and $2.20/M output via fireworks.

What are the context window sizes for DeepSeek-V3.2 (Thinking) and GLM-4.7?

DeepSeek-V3.2 (Thinking) supports 131K tokens and GLM-4.7 supports 203K tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between DeepSeek-V3.2 (Thinking) and GLM-4.7?

Key differences include context window (131K vs 203K), input pricing ($0.28 vs $0.60/M), multimodal support (no vs yes). See the full comparison above for benchmark-by-benchmark results.

Who makes DeepSeek-V3.2 (Thinking) and GLM-4.7?

DeepSeek-V3.2 (Thinking) is developed by DeepSeek and GLM-4.7 is developed by Zhipu AI.