Shipped · API staged

GLM-5.3
Same base, post-training only

As of 2026-10-20: Z.ai's official release notes date GLM-5.3 to 2026-08-18 and GLM-5.3-Flash to 2026-08-26; the Z.ai blog announcement is dated 2026-08-14 (the blog page was empty on this check, so not re-verified). Z.ai says it uses the same base as GLM-5.2, with every gain from post-training. At launch Z.ai said the Coding Plan came first, with the standalone API and full weights staged; as of 2026-10-20 the standalone API is live and the weights are public on Hugging Face (zai-org/GLM-5.3).

Updated 2026-10-20

One key across Claude, GPT and Chinese models, billed per token.

#glm-5.3#post-training#Coding Plan#weights ~2 weeks

At a glance

14 Aug

Official post

z.ai/blog/glm-5.3. Opening line: Scaling post-training is all we did for GLM-5.3.

Same base

Versus 5.2

Officially the GLM-5.2 base — no new pretrain. Community still cites the ~744B 5.2 shape.

Staged

Who can use it

At launch: Coding Plan first, standalone API “coming soon,” full weights in about two weeks. As of 2026-10-20: the standalone API is live and the weights are public on Hugging Face.

low / high / max

Thinking levels

Three effort levels in the integration note. OpenCode says Go has 5.3, 1M context, same price as 5.2 — that is OpenCode’s rate, not a Z.ai public tariff.

What 5.3 actually changed

Official docs: GLM-5.3 currently supports text-only inputs, with a 1M-token context window and a maximum output length of 128K tokens. Reasoning is always on, with three reasoning effort levels, low, high and max, and disabling reasoning is no longer supported. If your app sets thinking.type to "disabled", change it to enabled, set reasoning_effort to low and switch the model ID to glm-5.3, otherwise the request fails. Not a new base model. Z.ai says the jump came from more environments, more diverse tasks, and more compute on post-training. The post stresses harder coding, long-horizon work, and cybersecurity. Those cyber skills are dual-use, so release is staged: controlled partners first, broader API later, weights last.

What people are watching

Official release notes: GLM-5.3 on 2026-08-18, where Z.ai claims a 50% gain over GLM-5.2 on Z.ai Code Bench; GLM-5.3-Flash on 2026-08-26, which Z.ai says matches Mythos 5 in white-box code review and vulnerability discovery. OpenCode wrote “GLM-5.3 now available in Go”, 1M context, same price as 5.2. Interconnects followed: coding plan now, API soon, Hugging Face in two weeks. Vendor tables show jumps versus Kimi K3, Fable 5 and GPT-5.6 Sol. Independent reruns are not there yet, so do not treat those rows as a ranking.

Timeline

16 Jun 2026

GLM-5.2 ships and becomes the open-weight flagship of the line.

14 Aug 2026

z.ai/blog announces 5.3 (earlier record, not re-verified); the official release notes also list 2026-08-18. Z.ai says it is available to all GLM Coding Plan users. OpenCode Go lists it the same day.

~two weeks after launch

Official promise: publish full weights after safety evaluation. Standalone API still “coming soon.”

Confirmed vs treat carefully

Confirmed

Official release notes list GLM-5.3 and GLM-5.3-Flash; same base as 5.2; Coding Plan first, and Z.ai says it is available to all GLM Coding Plan users; API and weights staged; the official account states cyber dual-use is why the broad release waits. Thinking levels low / high / max are in the official integration note.

Treat carefully

Do not promote vendor Terminal-Bench / DeepSWE jumps as independent ranks. “Weights in two weeks” is not a download link today. Do not assume every reseller catalog added glm-5.3 on launch day.

Stay on 5.2 or chase 5.3

Keep GLM-5.2

You need MIT weights, a public API, and a Zhipu SKU you can order from a multi-model catalog today.

Wait / use a 5.3 channel

You already sit on Z.ai Coding Plan or OpenCode Go and can live with a staggered API and weights.

How to reach it

The official path is Z.ai Coding Plan. The standalone API id is previewed as glm-5.3; price and GA date are not locked on Z.ai’s public tariff. OpenCode users should read Go’s live model list. Local weights wait for a Hugging Face card — a promise is not a URL.

On QCode

QCode added glm-5.3 in Aug 2026; one key calls both glm-5.3 and glm-5.3-flash. The Zhipu line is now glm-5.3-flash, glm-5.3 and glm-5.2. Billed per token; per-model rates are on /models.

GLM-5.3 FAQ

Did 5.3 actually ship?

Yes. The Z.ai blog announcement was dated 2026-08-14 (earlier record, not re-verified); the official release notes date GLM-5.3 to 2026-08-18 and GLM-5.3-Flash to 2026-08-26. That is not the same as every API or aggregator listing it.

When can I use GLM-5.3, and can I download the weights?

Z.ai says GLM-5.3 is available to all GLM Coding Plan users, and the official price table now lists GLM-5.3. For the full weights, the official wording at launch was only that they follow the safety evaluation, with no fixed date; the official docs fetched on 2026-10-20 give no weights date either, and the latest status is on the GLM-5.3 status tracker page.

What is different from 5.2?

Same base, no new pretrain. Officially the jump is post-training. Treat vendor benches as vendor benches until someone else reruns them.

Can I call 5.3 on QCode right now?

Yes. The catalog lists glm-5.3-flash, glm-5.3 and glm-5.2.

What is the price?

Z.ai's official price table now lists GLM-5.3: $1.4 input, $0.26 cached input, $4.4 output per 1M tokens; GLM-5.2 is the same. OpenCode's “same as 5.2” is OpenCode's own rate, not the Z.ai price table.

Does it have native vision?

GLM-5.3 itself currently supports text-only inputs. Native vision arrives with GLM-5.3-Flash, released on 2026-08-26. The launch post sells coding, long-horizon agents and cyber; it does not pitch native vision as a 5.3 feature. Drop the pre-launch “5.3 will be multimodal” rumor.

Sources

z.ai/blog/glm-5.3; Z.ai's official GLM-5.3 docs and pricing pages; Z.ai's official release notes: 2026-08-18 GLM-5.3, 2026-08-26 GLM-5.3-Flash; Z.ai's official account on staged release and cyber dual-use; write-ups from OpenCode and Interconnects. All pages checked on 2026-10-20.

Use the Zhipu SKU that is actually listed

Run the new model for a few days first: not ready for a monthly plan? Pay-as-you-go Packs start at $4 for a one-time quota you use within 7 days; already on a Personal Plan and short on quota? Buy a booster pack. GLM-5.2 is on the catalog, and 5.3 is callable too.

Not sure where to start?

Try it for a week

Small Pack: $4 for $30 of quota, valid 7 days

Use it every day

Starter: $9.00/month with $5 of quota every day — just $0.060 per $1 of quota, vs about $0.113 for the Medium Pack