BroadRouter

GLM 5 vs GLM 4.7 API pricing

Updated

GLM 5: USD 0.60 input, USD 1.92 output per 1M tokens (GMI Cloud, promotional). GLM 4.7: USD 0.40 input, USD 1.75 output per 1M tokens (DeepInfra). At 3 input tokens per output token, GLM 5 costs 26% more than GLM 4.7 (USD 0.93 vs USD 0.7375 per 1M).

GLM 4.7 is the release before GLM 5 in the same Z.ai line (by version in the name), so this page shows what the newer release costs. Price only: every number is computed from prices read from each provider's own pricing page (linked per cell, newest check 2026-11-15), with no claim about quality, speed or accuracy. Third-party benchmark results both models share are quoted as published.

How do GLM 5 and GLM 4.7 differ on price?

GLM 5 and GLM 4.7 prices side by side

USD per 1M tokens, standard tier. "Cheapest" is the lowest price at 3 input tokens to 1 output token; each line can come from a different provider.

GLM 5 and GLM 4.7 API prices compared
PriceGLM 5GLM 4.7
Model pageGLM 5
Z.ai · 10 providers
GLM 4.7
Z.ai · 8 providers
Cheapest standard price, blended 3:1USD 0.93 promo
USD 0.60 in / USD 1.92 out at GMI Cloud
GMI Cloud page checked 2026-11-15
USD 0.7375 lower
USD 0.40 in / USD 1.75 out at DeepInfra
DeepInfra page checked 2026-11-14
Cheapest inputUSD 0.60 promo
GMI Cloud source
USD 0.40 lower
DeepInfra source
Cheapest cached inputUSD 0.10
Google Vertex AI source
USD 0.06 lower
Google Vertex AI source
Cheapest outputUSD 1.92 promo
GMI Cloud source
USD 1.75 lower
DeepInfra source
Cheapest non-promotional priceUSD 0.95 / USD 2.55
SiliconFlow, USD 1.35 blended source
The cheapest price is not promotional.
Maker's own APIUSD 1 / USD 3.20
Z.ai source
USD 0.60 / USD 2.20
Z.ai source
Median standard priceUSD 1 / USD 3.20
input / output, 10 prices
USD 0.60 / USD 2.20
input / output, 8 prices
Context window204,800 tokens, largest stated; from 198,000 at some providers204,800 tokens, largest stated; from 198,000 at some providers
Providers with a standard price108

Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).

What do GLM 5 and GLM 4.7 cost for a real workload?

30 days at each model's cheapest standard price for the workload (presets of the cost calculator; cached input at the cached price where listed).

Monthly cost of three workloads on GLM 5 and GLM 4.7
WorkloadGLM 5GLM 4.7Difference
Support chatbot
2,000 requests a day, 1,500 input tokens (60% cached), 300 output
USD 62.64
GMI Cloud
USD 50.22
DeepInfra
USD 12.42
GLM 4.7 lower
RAG over documents
1,000 requests a day, 8,000 input tokens (30% cached), 500 output
USD 138.24
GMI Cloud
USD 99.21
DeepInfra
USD 39.03
GLM 4.7 lower
Coding agent (one developer)
1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning
USD 518.40
GMI Cloud
USD 387.90
DeepInfra
USD 130.50
GLM 4.7 lower

GLM 4.7 is cheaper on all three workloads. The widest gap is on RAG over documents: the dearer model costs 39% more than the cheaper one.

Does the cheaper model change with your input:output ratio?

GLM 4.7 is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below GLM 5's, so the input:output ratio never changes the answer.

Price per 1M tokens of GLM 5 and GLM 4.7 at different input:output mixes
Input : output tokensGLM 5GLM 4.7
1 : 3 (long answers)USD 1.59
GMI Cloud
USD 1.413 lower
DeepInfra
1 : 1USD 1.26
GMI Cloud
USD 1.075 lower
DeepInfra
3 : 1 (the site’s blend)USD 0.93
GMI Cloud
USD 0.7375 lower
DeepInfra
10 : 1 (retrieval, long prompts)USD 0.72
GMI Cloud
USD 0.5227 lower
DeepInfra
30 : 1 (agents, big contexts)USD 0.6426
GMI Cloud
USD 0.4435 lower
DeepInfra

USD per 1M tokens at the cheapest standard provider for each mix, before cache discounts. Thinking tokens bill as output.

What do GLM 5 and GLM 4.7 cost per character in other languages?

Not measured: the real cost by language dataset covers only tokenizers we could measure from published files, and GLM 4.7 is not among them. Count tokens on your own text with the counters below.

Published results on the same benchmark

BroadNet did not run these benchmarks. BroadRouter is BroadNet's site. Each row pairs results from one source with the same setting for both models, as published; results from different sources or settings are never put side by side. One benchmark says little about your own workload.

Published results for GLM 5 and GLM 4.7 on the same benchmark, with sources
BenchmarkGLM 5GLM 4.7Source and setting
GPQA-Diamond86.085.7Z.ai model card: GLM 5, GLM 4.7, checked 2026-11-16
Setting not stated by the source
SWE-bench Verified77.873.8Z.ai model card: GLM 5, GLM 4.7, checked 2026-11-16
Setting not stated by the source
GPQA Diamond0.8780.833Epoch AI, checked 2026-11-16
Setting as listed by Epoch (no suffix), mean score on a 0 to 1 scale

Epoch AI rows: Epoch AI, Capabilities & benchmarking, licensed CC BY 4.0, shown as the mean score on a 0 to 1 scale, rounded to three decimals (hover a value for the exact figure).

All published results for each model, with every setting: GLM 5 · GLM 4.7.

Check it on your own traffic

More comparisons: GLM 5.1 vs GLM 5 · GLM 4.7 vs GLM 4.6 · GLM 4.7 Flash vs GLM 4.7 · every GLM 5 comparison · every GLM 4.7 comparison · all 662 price comparisons

Questions

Is GLM 5 cheaper than GLM 4.7?

No. On a blend of 3 input tokens to 1 output token, GLM 4.7 costs USD 0.7375 per 1M tokens and GLM 5 USD 0.93, at each model's cheapest standard price. For input alone, GLM 4.7 is cheaper (USD 0.40 against USD 0.60 per 1M); for output alone, GLM 4.7 is (USD 1.75 against USD 1.92). This compares price only, not quality.

What do GLM 5 and GLM 4.7 cost per 1M tokens?

GLM 5: from USD 0.60 input and USD 1.92 output at GMI Cloud (promotional); Z.ai's own API charges USD 1 / USD 3.20; median of 10 standard prices USD 1 / USD 3.20. GLM 4.7: from USD 0.40 input and USD 1.75 output at DeepInfra; Z.ai's own API charges USD 0.60 / USD 2.20; median of 8 standard prices USD 0.60 / USD 2.20. Each price is read from the provider's own pricing page; the source is linked in the table.

Which providers sell both GLM 5 and GLM 4.7?

7 providers have a standard price for both: GMI Cloud (USD 0.93 and USD 1 at 3:1), Featherless (USD 1.50 and USD 0.9625 at 3:1), Amazon Bedrock (USD 1.55 and USD 1 at 3:1), Google Vertex AI (USD 1.55 and USD 1 at 3:1), Novita AI (USD 1.55 and USD 1 at 3:1), Z.ai (USD 1.55 and USD 1 at 3:1), Venice (USD 1.55 and USD 1.075 at 3:1). The first is the lowest combined price. Buying each model from its own cheapest provider costs USD 0.93 and USD 0.7375.

Which is cheaper for a coding agent or a chatbot, GLM 5 or GLM 4.7?

For a coding agent (1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning): USD 518.40 a month on GLM 5 (GMI Cloud) against USD 387.90 on GLM 4.7 (DeepInfra), a difference of USD 130.50. For a support chatbot (2,000 requests a day, 1,500 input tokens (60% cached), 300 output): USD 62.64 a month on GLM 5 (GMI Cloud) against USD 50.22 on GLM 4.7 (DeepInfra), a difference of USD 12.42. Each uses the cheapest standard price for that workload, cached input included. Try your own numbers.

At what input:output ratio does the cheaper of GLM 5 and GLM 4.7 change?

GLM 4.7 is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below GLM 5's, so the input:output ratio never changes the answer.

Do GLM 5 and GLM 4.7 have cached input pricing?

GLM 5 from USD 0.10 per 1M cached input tokens (Google Vertex AI; 9 of 10 standard prices list a discount). GLM 4.7 from USD 0.06 per 1M cached input tokens (Google Vertex AI; 7 of 8 standard prices list a discount). Cached tokens are billed at the input price where no cache price is published.

Which has the larger context window, GLM 5 or GLM 4.7?

Both go up to 204,800 tokens on the providers that state it.

Which scores higher on GPQA-Diamond, GLM 5 or GLM 4.7?

Per the Z.ai model cards (GLM 5, GLM 4.7) (checked 2026-11-16), GLM 5 scores 86.0 and GLM 4.7 scores 85.7 on GPQA-Diamond (setting not stated by the source). BroadNet did not run these benchmarks, and one benchmark says little about your own workload. Every result both models share, with sources.

Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity