DeepSeek V4 Pro vs GLM-5.2: API Pricing & Benchmarks
DeepSeek V4 Pro vs GLM-5.2 — or, searched the other way round, GLM-5.2 vs DeepSeek V4 Pro: official API prices, what the same workload costs on each, and every published benchmark score, side by side. Every figure comes straight from the data below; this page does not pick a winner.
Price per 1M tokens
| DeepSeek V4 Pro | GLM-5.2 | |
|---|---|---|
| Input | $1.32 | $1.1952 |
| Output | $3.96 | $4.1831 |
| Blended | $1.98 | $1.9421 |
| Developer | DeepSeek | Zhipu AI |
| Released | 2026-04-24 | 2026-06-13 |
DeepSeek V4 Pro: Off-peak rate; peak-hour pricing per the official site; Single public source; verify on the official site
GLM-5.2: Compare the two price sources — models.dev repository, docs.z.ai (cited by LiteLLM)
GLM-5.2: Sources disagree; check the official site
Prices: official list prices as recorded by litellm (DeepSeek V4 Pro) and models.dev (GLM-5.2).
Cost of the same workload
| Workload | Tokens | DeepSeek V4 Pro | GLM-5.2 | Difference |
|---|---|---|---|---|
| Short chat reply | 500 in / 300 out | 0.185¢ | 0.185¢ | DeepSeek V4 Pro cheaper by 0.00045¢ |
| RAG query | 4,000 in / 500 out | 0.726¢ | 0.687¢ | GLM-5.2 cheaper by 0.0388¢ |
| Long document summary | 50,000 in / 1,000 out | $0.0700 | $0.0639 | GLM-5.2 cheaper by 0.602¢ |
| Code review of a 2k-line file | 30,000 in / 2,000 out | $0.0475 | $0.0442 | GLM-5.2 cheaper by 0.33¢ |
| 1,000 chat replies/day for 30 days | 30,000 × (500 in / 300 out) | $55.44 | $55.57 | DeepSeek V4 Pro cheaper by $0.1349 |
Cost = (input tokens ÷ 1,000,000 × input price) + (output tokens ÷ 1,000,000 × output price), using each model's own prices above.
Benchmarks
| Benchmark | DeepSeek V4 Pro | GLM-5.2 |
|---|---|---|
| Epoch Capabilities Index (ECI) | 150.1 epoch · as of 2026-04-24 |
153.0 epoch · as of 2026-06-16 |
| Epoch Mock AIME accuracy | 96.7% epoch · as of 2026-04-24 |
86.4% epoch · as of 2026-06-16 |
| Epoch FrontierMath (T1–3) accuracy | 45.3% epoch · as of 2026-04-24 |
59.2% epoch · as of 2026-06-16 |
| Arena text rating | 1458 arena · as of 2026-10-28 |
1475 arena · as of 2026-10-28 |
| Arena coding rating (human preference votes, not task accuracy) | 1500 arena · as of 2026-10-28 |
1514 arena · as of 2026-10-28 |
| Arena math rating | 1444 arena · as of 2026-10-28 |
1487 arena · as of 2026-10-28 |
| Arena WebDev rating | 1447 arena · as of 2026-10-28 |
1606 arena · as of 2026-10-28 |
Full pages
DeepSeek V4 Pro pricing GLM-5.2 pricing DeepSeek V4 Pro token counter GLM-5.2 token counter
FAQ
Which is cheaper, DeepSeek V4 Pro or GLM-5.2?
GLM-5.2 is cheaper on blended price: $1.9421 vs $1.98 per 1M tokens, which is $0.0379 less per 1M tokens (1.9% below DeepSeek V4 Pro).
How do the input and output prices of DeepSeek V4 Pro and GLM-5.2 compare?
DeepSeek V4 Pro costs $1.32 per 1M input tokens and $3.96 per 1M output tokens. GLM-5.2 costs $1.1952 per 1M input tokens and $4.1831 per 1M output tokens.
What are the ECI scores of DeepSeek V4 Pro and GLM-5.2?
DeepSeek V4 Pro scores 150.1 on the Epoch Capabilities Index (ECI); GLM-5.2 scores 153.0 on the Epoch Capabilities Index (ECI).
Where does the DeepSeek V4 Pro vs GLM-5.2 comparison data come from?
Prices come from models.dev and LiteLLM (open MIT-licensed catalogs); benchmark scores come from Epoch AI and LMArena (open CC BY 4.0 datasets). This page states the numbers side by side and does not pick an overall winner — see /data-sources/.