Cheaper overall
GLM 5
8% lower blended rate
Higher benchmark scores
GPT-5.4 Mini
leads on 6 of 9 figures
Bigger context window
GPT-5.4 Mini
400,000 vs 204,800 tokens
Cheaper cached input
GPT-5.4 Mini
$0.075 vs $0.20, 63% less
Where each one wins
GPT-5.4 Mini
- Cheaper input tokens $0.75 vs $1.00, 25% less
- Cheaper cached input $0.075 vs $0.20, 63% less
- Larger context window 400,000 tokens
- Higher graduate-level science (GPQA Diamond) 87.5 vs 82
- Higher instruction following (IFBench) 73.3 vs 72.3
- Higher long-context reasoning (AA-LCR) 77 vs 75.7
- Ahead on 3 more benchmark figures
- Offers Batch and Flex pricing not listed for GLM 5
GLM 5
- Cheaper output tokens $3.20 vs $4.50, 29% less
- Higher expert-exam performance (Humanity's Last Exam) 29.3 vs 28.1
- Higher support-agent performance (ฯยฒ-Bench Telecom) 98.2 vs 83.3
- Higher answer reliability (Omniscience Non-Hallucination) 64.7 vs 10.2
What four workloads cost
One run, and the same run a thousand times.
Caching is what decides it: over twenty calls GPT-5.4 Mini costs $0.1380 against $0.2048, a difference of $66.80 across a thousand loops.
| Workload | GPT-5.4 Mini | GLM 5 | ร1,000 | Difference |
|---|---|---|---|---|
| Chat turn 1,000 in / 500 out | $0.0030 | $0.0026 | $3.00 / $2.60 | GLM 5 is 13% cheaper |
| RAG answer 10,000 in / 800 out | $0.0111 | $0.0126 | $11.10 / $12.56 | GPT-5.4 Mini is 12% cheaper |
| Agent loop 32,000 in / 700 out ร 20 calls | $0.1380 | $0.2048 | $138.00 / $204.80 | GPT-5.4 Mini is 33% cheaper |
| Long document 400,000 in / 3,000 out | GLM 5's context window holds 204,800 tokens, so a 400,000-token prompt does not fit. | |||
Price per 1M tokens
Neither model changes its rate with prompt length, so these rates apply to every request.
| Context band | Price | GPT-5.4 Mini | GLM 5 | Difference |
|---|---|---|---|---|
| Any prompt length | Input | $0.75 | $1.00 | GPT-5.4 Mini is 25% cheaper |
| Any prompt length | Cached input | $0.075 | $0.20 | GPT-5.4 Mini is 63% cheaper |
| Any prompt length | Output | $4.50 | $3.20 | GLM 5 is 29% cheaper |
Blended: GPT-5.4 Mini $1.6875, GLM 5 $1.55 per 1M โ one rate at 3:1 input to output, for comparing two models at a glance.
GPT-5.4 Mini priced by OpenAI, GLM 5 by Z.ai. Standard tier, pay-as-you-go.
Other pricing tiers
Only GPT-5.4 Mini lists Batch and Flex, at up to 50% off its own standard rate; GLM 5 offers none of them.
Every tier either model sells is here, so this is the whole price sheet. A saving is measured against that model's own standard rate โ how we read tiers.
| Tier | GPT-5.4 Mini | GLM 5 | Cheaper |
|---|---|---|---|
| Batch only on GPT-5.4 Mini | $0.375 in / $2.25 out 50% off Standard | Not offered | — |
| Flex only on GPT-5.4 Mini | $0.375 in / $2.25 out 50% off Standard | Not offered | — |
Benchmarks
They share 9 figures: GPT-5.4 Mini leads on 6, GLM 5 on 3. The widest gap is 54.5 points, on answer reliability (Omniscience Non-Hallucination).
Graduate-level science
GPQA Diamond โ Graduate-level questions in biology, chemistry & physics
GPT-5.4 Mini ahead by 5.5
Expert-exam performance
Humanity's Last Exam โ Expert-level questions across many academic domains
GLM 5 ahead by 1.2
Instruction following
IFBench โ Precise following of detailed instructions
GPT-5.4 Mini ahead by 1
Support-agent performance
ฯยฒ-Bench Telecom โ Tool-using agent tasks in a telecom support setting
GLM 5 ahead by 14.9
Long-context reasoning
AA-LCR โ Long-context reasoning across large inputs
GPT-5.4 Mini ahead by 1.3
Command-line work
Terminal-Bench Hard โ Complex command-line and terminal workflows
GPT-5.4 Mini ahead by 9.1
Physics research reasoning
CritPt โ Unpublished physics research reasoning problems
GPT-5.4 Mini ahead by 8
Factual accuracy
Omniscience Accuracy โ Breadth of factual knowledge across domains
GPT-5.4 Mini ahead by 11.2
Answer reliability
Omniscience Non-Hallucination โ How reliably the model avoids fabricated answers
GLM 5 ahead by 54.5
Each figure is that model's best published run, with the effort level named beside it โ where these scores come from.
Specifications
GPT-5.4 Mini holds 195,200 more input tokens in one request. GPT-5.4 Mini also takes file and image.
| Specification | GPT-5.4 Mini | GLM 5 |
|---|---|---|
| Context window | 400,000 tokens | 204,800 tokens |
| Max output | 128,000 tokens | 128,000 tokens |
| Takes in | Text, image, file | Text |
| Puts out | Text | Text |
| Knowledge cutoff | 31 Aug 2025 | โ |
| Released | 17 Mar 2026 | 11 Feb 2026 |
| Status | Active | Active |
| Sold by | OpenAI, Perplexity | Z.ai |
FAQs
- Is GPT-5.4 Mini cheaper than GLM 5?
- No โ the other way round. A 1,000-token prompt with a 500-token reply costs $0.0030 on GPT-5.4 Mini and $0.0026 on GLM 5.
- How much do GPT-5.4 Mini and GLM 5 cost per 1M tokens?
- GPT-5.4 Mini costs $0.75 for input and $4.50 for output. GLM 5 costs $1.00 and $3.20. Standard pay-as-you-go rates, as of 11 Feb 2026.
- Which is better, GPT-5.4 Mini or GLM 5?
- On published benchmarks, GPT-5.4 Mini. It leads on 6 of the 9 figures both models report, including graduate-level science (GPQA Diamond), where it scores 87.5 against 82.
- Does GPT-5.4 Mini or GLM 5 offer batch pricing?
- GPT-5.4 Mini only. Its batch input costs $0.375, 50% off its standard rate, and GLM 5 lists no batch tier at all.
- Does GPT-5.4 Mini or GLM 5 have a bigger context window?
- GPT-5.4 Mini, at 400,000 tokens against 204,800. That is 195,200 more input tokens in a single request.