DeepSeek V4 vs GLM-5
Benchmarks, API pricing and specs, head to head. Data updated 2026-06-10.
DeepSeek V4
DeepSeek · Mar 2026
The open-weights shock of 2026: ~81% SWE-bench Verified, a 1M context and MIT license at $0.30/$0.50 per million tokens.
GLM-5
Z.ai (Zhipu) · Feb 2026
Z.ai's 744B open-weights MoE — within ~4 points of Opus 4.6 on SWE-bench Verified at $1/$3.20 per million tokens.
The verdict
DeepSeek V4 wins 1 of the 1 benchmarks these models share, against 0 for GLM-5. On price the gap is dramatic: DeepSeek V4 works out roughly 4.4x cheaper per blended million tokens. DeepSeek V4 also takes 1M of context versus 200K for GLM-5.
Benchmark head-to-head 1–0
Specs & pricing
| DeepSeek V4 | GLM-5 | |
|---|---|---|
| modhub Index | — | — |
| Input price / 1M | $0.3 | $1 |
| Output price / 1M | $0.5 | $3.2 |
| Context window | 1M | 200K |
| Max output | 64K | 128K |
| Open weights | yes (MIT) | yes (MIT) |
| Reasoning model | yes | yes |
| Multimodal input | text | text |
| Knowledge cutoff | Dec 2025 | Dec 2025 |
| Released | Mar 2026 | Feb 2026 |
| Example monthly cost* | $3.75 | $14.80 |
* 10M input + 1.5M output tokens per month at list prices, no caching. Green = better value on that row.
Frequently asked questions
- Which is better, DeepSeek V4 or GLM-5?
- DeepSeek V4 wins 1 of the 1 benchmarks these models share, against 0 for GLM-5. On price the gap is dramatic: DeepSeek V4 works out roughly 4.4x cheaper per blended million tokens. DeepSeek V4 also takes 1M of context versus 200K for GLM-5.
- Which is cheaper, DeepSeek V4 or GLM-5?
- DeepSeek V4 costs $0.3/$0.5 per million input/output tokens, while GLM-5 costs $1/$3.2. For a typical workload of 10M input and 1.5M output tokens per month, that's $3.75 versus $14.80.
- Which model is better for coding, DeepSeek V4 or GLM-5?
- On SWE-bench Verified — the standard agentic-coding benchmark — DeepSeek V4 scores ~81% versus ~77.8% for GLM-5, making DeepSeek V4 the stronger pick for coding agents.