Cheaper overall
GPT-5.4 Nano
49% lower blended rate
Higher benchmark scores
GPT-5.4 Nano
leads on 10 of 10 figures
Bigger context window
GPT-5.4 Nano
400,000 vs 65,536 tokens
Cheaper cached input
GPT-5.4 Nano
$0.02 vs $0.11, 82% less
Where each one wins
GPT-5.4 Nano
- Cheaper input tokens $0.20 vs $0.60, 67% less
- Cheaper output tokens $1.25 vs $1.80, 31% less
- Cheaper cached input $0.02 vs $0.11, 82% less
- Cheaper on all four workloads
- Larger context window 400,000 tokens
- Longer maximum output 128,000 tokens
- Higher overall intelligence (Intelligence Index) 20.7 vs 7.6
- Higher graduate-level science (GPQA Diamond) 81.7 vs 68.4
- Higher expert-exam performance (Humanity's Last Exam) 28.3 vs 6.3
- Ahead on 7 more benchmark figures
- Offers Batch and Flex pricing not listed for GLM 4.5V
GLM 4.5V
- Not scheduled for retirement GPT-5.4 Nano retires 1 Apr 2027
What four workloads cost
One run, and the same run a thousand times.
GPT-5.4 Nano is cheaper on all four workloads, by much the same margin each time โ about 2.5 times, from $0.0008 against $0.0015 on a chat turn to $0.0375 against $0.1152 on a long document.
| Workload | GPT-5.4 Nano | GLM 4.5V | ร1,000 | Difference |
|---|---|---|---|---|
| Chat turn 1,000 in / 500 out | $0.0008 | $0.0015 | $0.8250 / $1.50 | GPT-5.4 Nano is 45% cheaper |
| RAG answer 10,000 in / 800 out | $0.0030 | $0.0074 | $3.00 / $7.44 | GPT-5.4 Nano is 60% cheaper |
| Agent loop 32,000 in / 700 out ร 20 calls | $0.0375 | $0.1152 | $37.50 / $115.20 | GPT-5.4 Nano is 67% cheaper |
| Long document 400,000 in / 3,000 out | GLM 4.5V's context window holds 65,536 tokens, so a 400,000-token prompt does not fit. | |||
Price per 1M tokens
Above 272,000 input tokens the rates change, and the new rate applies to the whole request, not just the tokens past that point. It does not change which of the two is cheaper.
| Context band | Price | GPT-5.4 Nano | GLM 4.5V | Difference |
|---|---|---|---|---|
| Prompts up to 272,000 tokens | Input | $0.20 | $0.60 | GPT-5.4 Nano is 67% cheaper |
| Prompts up to 272,000 tokens | Cached input | $0.02 | $0.11 | GPT-5.4 Nano is 82% cheaper |
| Prompts up to 272,000 tokens | Output | $1.25 | $1.80 | GPT-5.4 Nano is 31% cheaper |
| Prompts over 272,000 tokens | Input | $0.20 | $0.60 | GPT-5.4 Nano is 67% cheaper |
| Prompts over 272,000 tokens | Cached input | $0.02 | $0.11 | GPT-5.4 Nano is 82% cheaper |
| Prompts over 272,000 tokens | Output | $1.25 | $1.80 | GPT-5.4 Nano is 31% cheaper |
Blended: GPT-5.4 Nano $0.4625, GLM 4.5V $0.90 per 1M โ one rate at 3:1 input to output, for comparing two models at a glance.
GPT-5.4 Nano priced by OpenAI, GLM 4.5V by Z.ai. Standard tier, pay-as-you-go.
Other pricing tiers
Only GPT-5.4 Nano lists Batch and Flex, at up to 50% off its own standard rate; GLM 4.5V 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 Nano | GLM 4.5V | Cheaper |
|---|---|---|---|
| Batch only on GPT-5.4 Nano | $0.10 in / $0.625 out 50% off Standard | Not offered | — |
| Flex only on GPT-5.4 Nano | $0.10 in / $0.625 out 50% off Standard | Not offered | — |
Benchmarks
GPT-5.4 Nano leads on all 10 figures, furthest ahead on long-context reasoning (AA-LCR), by 76.7 points.
Overall intelligence
Intelligence Index โ Overall intelligence across reasoning, knowledge & math evals
GPT-5.4 Nano ahead by 13.1
Graduate-level science
GPQA Diamond โ Graduate-level questions in biology, chemistry & physics
GPT-5.4 Nano ahead by 13.3
Expert-exam performance
Humanity's Last Exam โ Expert-level questions across many academic domains
GPT-5.4 Nano ahead by 22
Instruction following
IFBench โ Precise following of detailed instructions
GPT-5.4 Nano ahead by 41.7
Support-agent performance
ฯยฒ-Bench Telecom โ Tool-using agent tasks in a telecom support setting
GPT-5.4 Nano ahead by 53.5
Long-context reasoning
AA-LCR โ Long-context reasoning across large inputs
GPT-5.4 Nano ahead by 76.7
Command-line work
Terminal-Bench Hard โ Complex command-line and terminal workflows
GPT-5.4 Nano ahead by 35.6
Physics research reasoning
CritPt โ Unpublished physics research reasoning problems
GPT-5.4 Nano ahead by 9.3
Factual accuracy
Omniscience Accuracy โ Breadth of factual knowledge across domains
GPT-5.4 Nano ahead by 4.9
Answer reliability
Omniscience Non-Hallucination โ How reliably the model avoids fabricated answers
GPT-5.4 Nano ahead by 33.7
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 Nano holds 334,464 more input tokens in one request. GPT-5.4 Nano also takes file, and GLM 4.5V also takes video. GPT-5.4 Nano retires on 1 Apr 2027, so anything built on it needs a move planned.
| Specification | GPT-5.4 Nano | GLM 4.5V |
|---|---|---|
| Context window | 400,000 tokens | 65,536 tokens |
| Max output | 128,000 tokens | 16,384 tokens |
| Takes in | Text, image, file | Text, image, video |
| Puts out | Text | Text |
| Knowledge cutoff | 31 Aug 2025 | 31 Dec 2024 |
| Released | 17 Mar 2026 | 11 Aug 2025 |
| Status | Legacy | Active |
| Retires | 1 Apr 2027 | โ |
| Sold by | OpenAI | Z.ai |
FAQs
- Is GPT-5.4 Nano cheaper than GLM 4.5V?
- Yes. A 1,000-token prompt with a 500-token reply costs $0.0008 on GPT-5.4 Nano and $0.0015 on GLM 4.5V, and the same model is cheaper on every workload on this page.
- How much do GPT-5.4 Nano and GLM 4.5V cost per 1M tokens?
- GPT-5.4 Nano costs $0.20 for input and $1.25 for output. GLM 4.5V costs $0.60 and $1.80. Standard pay-as-you-go rates, as of 11 Aug 2025.
- Which is better, GPT-5.4 Nano or GLM 4.5V?
- On published benchmarks, GPT-5.4 Nano. It leads on 10 of the 10 figures both models report, including overall intelligence (Intelligence Index), where it scores 20.7 against 7.6.
- Does GPT-5.4 Nano or GLM 4.5V offer batch pricing?
- GPT-5.4 Nano only. Its batch input costs $0.10, 50% off its standard rate, and GLM 4.5V lists no batch tier at all.
- Does GPT-5.4 Nano or GLM 4.5V have a bigger context window?
- GPT-5.4 Nano, at 400,000 tokens against 65,536. That is 334,464 more input tokens in a single request.