Zhipu · preview runs only
GLM 5.3 Flash and Taskade
GLM 5.3 Flash from Zhipu is not in the Taskade model menu. We ran it in preview tests, which the TSK-1 ranking does not count. Its newest preview run was on Oct 23, 2026. TSK-1, our app-building test, does not rank preview runs.
- Status
- Not in Taskade
- Context window
- 1MtokensHow much it can read at once
- Max output
- 131KtokensLongest answer it can write
- Released
- Where it runs
- 16servicesOther services that serve it
Published scores
Benchmark scores that Zhipu and test sites published for GLM 5.3 Flash. Taskade did not re-run them. Lab-reported marks a score from Zhipu itself.
| Benchmark | Score | Source |
|---|---|---|
| Terminal-Benchfirst try · v2.1 · max effort · in Claude Code 2.1.207Hugging FaceLab-reported | 84.3% | Hugging FaceLab-reported |
| Humanity's Last Examaccuracy · max effort · with tools · full setHugging FaceLab-reported | 55.3% | Hugging FaceLab-reported |
More published scores (4)
| Benchmark | Score | Source |
|---|---|---|
| DeepSWEresolved · v1.1 · max effort · in mini-swe-agentHugging FaceLab-reported | 63.4% | Hugging FaceLab-reported |
| Agents' Last Exammax effortHugging FaceLab-reported | 26.3 | Hugging FaceLab-reported |
| AutomationBenchfirst try · v1.0.6 · max effortHugging FaceLab-reported | 48.8% | Hugging FaceLab-reported |
| GDPval-AAElo rating · v2 · max effortHugging FaceLab-reported | 1,773 | Hugging FaceLab-reported |
LMArena scores
| Benchmark | Score | Source |
|---|---|---|
| Text arena±5 · 41,642 votes · Rank 33 of 419 · LMArena | 1,477 | LMArena |
| WebDev arena±6 · 21,587 votes · Rank 32 of 152 · LMArena | 1,607 | LMArena |
Scores from the LMArena leaderboard dataset (CC BY 4.0). People vote between two answers, and the votes set the score. Taskade picked the Overall category and the default setting of each model.
Scores collected by models.dev (MIT). Each source links the page that published the score.
Versions tested
Preview runs only, not ranked
Model facts
Published specs from the model maker. A token is about three quarters of a word. Taskade can run the model with lower limits than these.
- Last updated
- Reads
- Text, Images, Video, PDFWhat you can give it
- Writes
- TextWhat it gives back
- Reasoning
- YesThinking a problem through first
- Tool calling
- YesUsing tools, such as web search
- File attachments
- YesReading files that you attach
- Structured output
- YesFilling a set format, such as a table
- Open weights
- YesFree for anyone to download and run
- Thinking levels
- LowHighMaxHow long it can think. Higher is slower and more careful.
Published specs from the model maker, via models.dev (MIT).
Where it runs
Services outside Taskade that serve GLM 5.3 Flash, from models.dev. Each link opens in a new tab.
What to pick in Taskade
- Leave TSK-Auto on, and Taskade picks a current model for each request.
- To pick a model yourself, choose one from the model menu.
- To compare, open the models page and read the test results of the models you can pick.
Compare GLM
Other Zhipu models
FAQ
Can I pick GLM 5.3 Flash in Taskade?
No. GLM 5.3 Flash is not in the Taskade model menu, and the key providers in Taskade do not list it by name. TSK-Auto picks a current model for each request. The models page lists the models you can pick now.
Has Taskade tested GLM 5.3 Flash?
Only in preview runs, last on Oct 23, 2026. The TSK-1 ranking does not count them, so GLM 5.3 Flash has no TSK-1 score. The TSK-1 page shows what each run built.
Who makes GLM 5.3 Flash?
Zhipu makes GLM 5.3 Flash. Taskade runs frontier models from top AI labs, and the GLM results page shows how Zhipu models build apps.