Download

Claude Sonnet 5.5 vs GLM-5.3

GLM-5.3 is cheaper than Claude Sonnet 5.5 per token, $1.40 input and $4.40 output per 1M against $2 and $10, and its weights are open. Prompt caching closes most of the gap: our cached feature session costs $2.32 on GLM-5.3 and $2.59 on Sonnet 5.5, because cached input on GLM ($0.26) costs more than on Sonnet ($0.10).

Price with and without the cache

Uncached, our feature session costs $7.25 on GLM-5.3 and $10.80 on Sonnet 5.5, and a long agent day $66.30 against $97.50. With 90% of prompts read from the cache the long day is $20.13 against $20.55, nearly the same. A coding agent resends the same conversation with every call, which is the work the cache does, so the cached figures are closer to a real session.

Running GLM-5.3 in Claude Code

Z.ai runs an endpoint that speaks the Anthropic Messages API, so Claude Code can send its requests to GLM instead of Claude. Those requests are billed by Z.ai, not covered by a Claude plan, and Anthropic doesn't support that routing. Using Claude Code with GLM or Kimi has the settings and what still works.

What GLM-5.3 reads, and what we know of its scores

GLM-5.3 reads text only, up to 1.05M tokens, and writes up to 131K tokens in one reply. Sonnet 5.5 reads text and images, up to 1M tokens, and writes up to 128K.

Sonnet 5.5 scored 70.6% on Terminal-Bench 4.0; we found no GLM-5.3 score on that or our other two benchmarks. GLM-5.3 was first listed on the API on 18 August 2026, six weeks before Anthropic released Sonnet 5.5 on 28 September.

Side by side

Claude Sonnet 5.5 and GLM-5.3: specifications and prices
Claude Sonnet 5.5GLM-5.3
MakerAnthropicZ.ai
Released28 September 2026[1]18 August 2026[2] first API listing
Context window1M tokens[1]1.05M tokens[3] As served by the price's provider on OpenRouter.
Max output128K tokens[1]131K tokens[3] As served by the price's provider on OpenRouter.
Input price$2 per 1M tokens[1]$1.40 per 1M tokens[3]
Output price$10 per 1M tokens[1]$4.40 per 1M tokens[3]
Cached input$0.10 per 1M tokens[4]$0.26 per 1M tokens[3]
Long promptsOne priceOne price
Price changeNone announcedNone announced
Open weightsNo[1] Offered through the Claude API and cloud platforms; no weights are published.Yes[5] Custom licence.
Readstext, image[1]text[2]

Benchmarks

None of the benchmarks below has a published score for both models. We only show a score we can cite with its source and date; "Not published" means we found none for that model, not that it scored zero.

Published benchmark scores (higher is better)
Claude Sonnet 5.5GLM-5.3
Terminal-Bench 4.0Claude Sonnet 5.5: 70.6%70.6%GLM-5.3: not publishedFrontierCode 1.1 (Main)Claude Sonnet 5.5: 46.2%46.2%GLM-5.3: not publishedHumanity's Last Exam (with tools)Claude Sonnet 5.5: 64.5%64.5%GLM-5.3: not published
Benchmarks: Claude Sonnet 5.5 and GLM-5.3
BenchmarkClaude Sonnet 5.5GLM-5.3
Terminal-Bench 4.0Agentic coding tasks in a real terminal.70.6%[6]Not published
FrontierCode 1.1 (Main)Hard software engineering tasks.46.2%[6] Max effort.Not published
Humanity's Last Exam (with tools)Expert-level questions across many fields, tools allowed.64.5%[6]Not published

What a coding session costs

Our own estimate from the list prices above, for three sessions we define: a coding agent resends the conversation with every call, so prompts are large and grow. "Cached" assumes 90% of prompt tokens are read from the prompt cache and leaves out cache write fees. A long-prompt price applies when a call's prompt is over its threshold. Work out your own mix with the token cost calculator.

Feature session at list prices, US dollars (lower is cheaper)
Claude Sonnet 5.5GLM-5.3
No cachingClaude Sonnet 5.5: $10.80$10.80GLM-5.3: $7.25$7.2590% of prompt tokens from cacheClaude Sonnet 5.5: $2.59$2.59GLM-5.3: $2.32$2.32
Session costs: Claude Sonnet 5.5 and GLM-5.3
SessionClaude Sonnet 5.5Claude Sonnet 5.5 cachedGLM-5.3GLM-5.3 cached
Quick fix10 calls of 25K prompt and 1.5K output tokens$0.65$0.22$0.42$0.16
Feature session60 calls of 80K prompt and 2K output tokens$10.80$2.59$7.25$2.32
Long agent day300 calls of 150K prompt and 2.5K output tokens$97.50$20.55$66.30$20.13

Value for money and release gap

Claude Sonnet 5.5 is dated 6 weeks later (41 days; GLM-5.3: first API listing). Price per point divides our feature-session cost by a published score: lower means more score for the money. It is a rough lens, not a ranking.

Value for money: Claude Sonnet 5.5 and GLM-5.3
MeasureClaude Sonnet 5.5GLM-5.3
Input tokens per $1500K714K
Filling the whole context window once$2.00$1.47

Questions

Is GLM-5.3 cheaper than Claude Sonnet 5.5?

Per token, yes: $1.40 against $2 per 1M input tokens and $4.40 against $10 per 1M output. With 90% of prompts cached, our feature session costs $2.32 on GLM-5.3 and $2.59 on Sonnet 5.5.

Does GLM-5.3 have open weights?

Yes, according to the source we cite. The weights of Claude Sonnet 5.5 are not open.

Can GLM-5.3 read images?

Not by the source we cite: GLM-5.3 reads text only. Claude Sonnet 5.5 reads text and images.

Most compared

Any other two models: pick them in the lists at the top of the page.

More tools

Sources

  1. platform.claude.com: docs/en/models/sonnet-5-5/overview, read
  2. openrouter.ai: z-ai/glm-5.3, read
  3. openrouter.ai: z-ai/glm-5.3/providers, read
  4. platform.claude.com: docs/en/about-claude/pricing, read
  5. huggingface.co: zai-org/GLM-5.3, read
  6. anthropic.com: claude-sonnet-5-5, read

Numbers from the makers' own pages, checked by hand, last read . List prices in US dollars before tax. Spotted a mistake? office@shuriken.sh.

Made by Shuriken, the harness for coding agents. Its hot swapping moves your chats to another of your accounts or API keys when one nears its limit. Use only accounts you own and are entitled to use.

Try Shuriken on your own computer

Add your Claude, Codex and Gemini accounts once. Shuriken switches between them so your agents keep working.

Download Pricing
Download on the App StoreGet it on Google PlayComing soon